24
Journal of Economic Literature, Vol. XXXIII ( December 1995), pp. 1879–1902 Dasgupta: The Population Problem The Population Problem: Theory and Evidence By PARTHA DASGUPTA University of Cambridge and Beijer International Institute of Ecological Economics, Stockholm An early version of this article was prepared as a background paper for the Population Summit of the World’s Scientific Academies, held in New Delhi during 24–27 October 1993, and was also the basis of the 1993 Sidney Ball Lecture at the University of Oxford, a much-shortened version of which was published in Scientific American (Dasgupta 1995a). These essays, in turn, were based on ideas that were developed in my book, An Inquiry into Well-Being and Destitution (Dasgupta 1993). In preparing this article, I have benefited greatly from discussions with Kenneth Arrow, Carol Dasgupta, Jack Goody, and Paul Seabright. I. Introduction: Motivation and Scope A S WITH POLITICS, we all have widely differing opinions about population. Some point to population growth as the cause of poverty and envi- ronmental degradation. Others permute the elements of this causal chain, argu- ing, for example, that poverty is the cause rather than the consequence of increasing numbers. Yet even when studying the semi-arid regions of sub-Sa- haran Africa and the Indian subconti- nent, economists have usually not re- garded population growth, poverty, and the state of the local environmental re- source-base as interconnected. Inquiry into each factor has in large measure gone along its own narrow route, with discussions of their interactions domi- nated by popular writings (e.g., Paul and Anne Ehrlich 1990), which, while often il- luminating, are in the main descriptive, not analytical. 1 Over the past several years, though, a few investigators have studied the inter- actions between the three ingredients more closely. Our approach has been to fuse theoretical modeling with empirical findings drawn from a number of disci- plines. The resulting construction re- gards none of the three factors to be the prior cause of the other two; rather, it sees each as influencing the others and in turn being influenced by them. Focus- ing on people in small, rural communi- ties in the poorest regions of the world, the work has even identified circum- stances in which population growth, pov- 1879 1 Nancy Birdsall (1988), Allen Kelley (1988), and T. Paul Schultz (1988) are three authoritative surveys on the subject of population growth in poor countries. None of them focuses on the inter- actions I am concerned with in this article. Nor does a large literature on poverty (e.g., Jean Drèze and Amartya Sen 1990) address our theme.

The Population Problem: Theory and Evidence · 2012-06-09 · The Population Problem: Theory and Evidence By PARTHA DASGUPTA University of Cambridge and Beijer International Institute

  • Upload
    others

  • View
    2

  • Download
    0

Embed Size (px)

Citation preview

Page 1: The Population Problem: Theory and Evidence · 2012-06-09 · The Population Problem: Theory and Evidence By PARTHA DASGUPTA University of Cambridge and Beijer International Institute

Journal of Economic Literature,Vol. XXXIII (December 1995), pp. 1879–1902

Dasgupta: The Population Problem

The Population Problem: Theory and Evidence

By PARTHA DASGUPTAUniversity of Cambridge

andBeijer International Institute of Ecological Economics, Stockholm

An early version of this article was prepared as a background paper for the PopulationSummit of the World’s Scientific Academies, held in New Delhi during 24–27 October 1993,and was also the basis of the 1993 Sidney Ball Lecture at the University of Oxford, amuch-shortened version of which was published in Scientific American (Dasgupta 1995a).These essays, in turn, were based on ideas that were developed in my book, An Inquiry intoWell-Being and Destitution (Dasgupta 1993). In preparing this article, I have benefitedgreatly from discussions with Kenneth Arrow, Carol Dasgupta, Jack Goody, and PaulSeabright.

I. Introduction: Motivation and Scope

A S WITH POLITICS, we all havewidely differing opinions about

population. Some point to populationgrowth as the cause of poverty and envi-ronmental degradation. Others permutethe elements of this causal chain, argu-ing, for example, that poverty is thecause rather than the consequence ofincreasing numbers. Yet even whenstudying the semi-arid regions of sub-Sa-haran Africa and the Indian subconti-nent, economists have usually not re-garded population growth, poverty, andthe state of the local environmental re-source-base as interconnected. Inquiryinto each factor has in large measuregone along its own narrow route, withdiscussions of their interactions domi-nated by popular writings (e.g., Paul andAnne Ehrlich 1990), which, while often il-

luminating, are in the main descriptive,not analytical.1

Over the past several years, though, afew investigators have studied the inter-actions between the three ingredientsmore closely. Our approach has been tofuse theoretical modeling with empiricalfindings drawn from a number of disci-plines. The resulting construction re-gards none of the three factors to be theprior cause of the other two; rather, itsees each as influencing the others andin turn being influenced by them. Focus-ing on people in small, rural communi-ties in the poorest regions of the world,the work has even identified circum-stances in which population growth, pov-

1879

1 Nancy Birdsall (1988), Allen Kelley (1988),and T. Paul Schultz (1988) are three authoritativesurveys on the subject of population growth inpoor countries. None of them focuses on the inter-actions I am concerned with in this article. Nordoes a large literature on poverty (e.g., Jean Drèzeand Amartya Sen 1990) address our theme.

Page 2: The Population Problem: Theory and Evidence · 2012-06-09 · The Population Problem: Theory and Evidence By PARTHA DASGUPTA University of Cambridge and Beijer International Institute

erty, and degradation of the local envi-ronmental resource-base can fuel oneanother over extended periods of time.In this article I will present an outline ofthis work, laying stress on the argumentsthat have shaped it (Sections III-VII)and on the policy recommendations thathave emerged from it (Section VIII). Aswe will see, this body of work offerssomething like a new perspective onwhat may be called the population prob-lem.

A decade ago an inquiry was made intothe economic consequences of popula-tion growth in poor countries (NationalResearch Council 1986). Drawing ontime-series and cross-regional data, theinvestigators observed, among otherthings, that population size and itsgrowth can have both positive and nega-tive effects. The investigators recognizedthat population growth should not be re-garded as exogenously given. Neverthe-less, for tractability, they treated it as acausal factor in their inquiry, and con-cluded that, while economic develop-ment in most poor countries would befaster with slower rates of populationgrowth, there is no cause for alarm overthe high rates being experienced there.2

That population growth is best re-garded as endogenous is the hallmark ofa strand in demographic theory—oftencalled the “new economic demography”(Schultz 1988)—that had been devel-oped earlier.3 It contains an elaborate ac-count of the determinants of fertility be-havior. One weakness of this literature isthat, with relatively few exceptions, ithas focused on decisions made by a sin-gle household; it has not studied in detailsocial mechanisms in which a myriad ofindividual household decisions lead tooutcomes that are a collective failure. In

short, the literature has often equatedprivate and social benefits (and costs) ofhaving children. (“The next step is to ap-ply . . . microeconomic models (of house-hold behavior) to understand aggregatedevelopments in a general equilibriumframework. But progress in this field hasbeen slow”; Schultz 1988, p. 418.) Thenew perspective, in contrast, focuses onvarious types of externalities that couldbe associated with fertility decisions. Itnotes that a number of such externalitiescan, over time, lead to wide divergencesbetween individual intentions and socialrealizations (Sections IV-VII). The the-ory also peers more closely into the char-acter of decision making within ruralhouseholds and relates fertility behaviorto gender issues (Section III).

To anyone who is not a demographer,economic demography can be a mostfrustrating subject. It would seem thatfor any theoretical inference (say, on fer-tility matters), no matter how innocuous,there is some set of data from some partof the world over some period that is notconsonant with it. Over 40 years of demo-graphic research have uncovered that thefactors underlying fertility behavior in-clude not only the techniques that areavailable to households for controllingtheir size (Section IV), but also the house-hold demand for children (Sections V-VI). The latter in particular is influenced bya number of factors, whose relativestrengths would be expected to differacross cultures, and over time within agiven culture, responsive as they are tochanges in income and wealth and thestructure of relative prices. Thus, thefactors that would influence the drop inthe total fertility rate in a society from,say, 7 to 5 should be expected to be dif-ferent from those that would influence thedrop from 5 to 3 in that same society.4

2 Kelley (1988) is an excellent review of thefindings.

3 Gary Becker (1981) is often regarded as thecanonical formulation.

4 Total fertility rate is the number of live birthsa woman would expect to give if she were to livethrough her child-bearing years and to bear chil-

1880 Journal of Economic Literature, Vol. XXXIII (December 1995)

Page 3: The Population Problem: Theory and Evidence · 2012-06-09 · The Population Problem: Theory and Evidence By PARTHA DASGUPTA University of Cambridge and Beijer International Institute

Across societies the matter is stillmore difficult. The springs of human be-havior in an activity at once so personaland social as procreation are complexand interconnected, and empirical con-firmation of ideas are always fraughtwith difficulty.

These observations are merely a re-minder that a general theory of fertilitybehavior is not currently available. Thefindings I will be reporting here enableus to weave a narrative (Section VII) ofthe circumstances in which some peopleare born, what they aspire to, and themanner in which they live and die. Thefindings cannot be applied universally.But there are reasons for thinking thatit may have a bearing in parts of the In-dian subcontinent and sub-Saharan Af-rica. The account offers us an interpreta-tion of the population problem in someof the poorest regions of the world. Itsees the problem as consisting of unac-ceptable risks of maternal death forpoor, illiterate women, and of new livesdoomed to extreme poverty amid dete-rioration of the local environmental re-source-base. In this sense, the new per-spective regards current rates ofpopulation growth in these regions to beoverly high and, so, a matter for concern.This is in marked contrast to the viewsexpressed in both the strands of thedemographic literature I have just al-luded to.

II. Population, GNP, Food Production,and the Environment

Why did demographic writings a dec-ade ago not regard high rates of popula-tion growth in poor countries a worryingmatter? In particular, why did investiga-

tors arrive at what Kelley (1988, p. 1686)calls a “revisionist interpretation” of theevidence?5

It seems to me there were at least tworeasons. One is the preoccupation ofthose who developed the new householdeconomics. Attention was paid in greatpart to choices made by a single, opti-mizing household; and the study of a sin-gle household is not a propitious one inwhich to explore the possibilities of col-lective failure.

A second reason stems from empiricalfindings. As the accompanying figuresuggests, richer countries on averagehave been associated with lower fertilityrates. A regional breakdown of even theChinese experience displays the generalpattern: fertility is lower in higher-in-come regions (Birdsall and Dean Jami-son 1983). Furthermore, barring sub-Sa-haran Africa over the past 25 years or so,gross income per head has grown innearly all poor regions since the end ofthe Second World War. In addition,growth in world food production since1960 has exceeded the world’s popula-tion growth by an annual rate of, ap-proximately, 0.6 percent. This has beenaccompanied by improvements in a num-ber of indicators of human well-being,such as the infant survival rate, life ex-pectancy at birth, and literacy. In poorregions all this has occurred in a regimeof population growth rates substantiallyhigher than in the past: excepting forEast Asia and parts of South and South-east Asia, modern-day declines in mor-tality rates have not been matched by re-ductions in fertility. A number of placesthat did experience a decline in fertilityrates for a while (e.g., Costa Rica and

dren at each age in accordance with the prevailingage-specific fertility rates. The measure pertainsto the number of live births, not pregnancies.

5 The “revision” pertains to a decided shift froman earlier view, expressed, for example, in the1971 National Academy of Sciences report entitledRapid Population Growth: Consequences and Pol-icy Implications, which found little good in popu-lation growth.

Dasgupta: The Population Problem 1881

Page 4: The Population Problem: Theory and Evidence · 2012-06-09 · The Population Problem: Theory and Evidence By PARTHA DASGUPTA University of Cambridge and Beijer International Institute

India), appear to have stabilized at levelswell above the population replacementrate (the fertility rate at which popula-tion size would be expected to stabilizein the long run; a figure just over 2.1).6

Table 1 presents (total) fertility ratesin several countries and groups of coun-tries. Toward the end of the 1980s, thefertility rate in the World Bank’s list oflow-income countries (excluding Chinaand India) was 5.6. The figures for Chinaand India were 2.3 and 4.2, respectively.(In contrast, fertility rates in western in-dustrial countries today lie between 1.5and 1.9.) The oft-expressed fear thatrapid population growth will accompany

deteriorations in living standards has notbeen borne out by experience so far; atleast, not when judged from the vantageof the world as a whole. Taken together,these observations suggest not only thatpopulation growth is not a serious hin-drance to economic growth, but that theprocess of economic growth itself can berelied upon to reduce the rate of popula-tion growth. This provides a reasoningbehind the revisionist interpretation.

Tota

l fer

tility

rate

6

0

5

1

4

3

2

Figure 1. Fertility in relation to income in developing countries, 1982

Source: Birdsall (1988)

7

8

$1000 $2000 $3000 $4000 $5000 $6000

Income per capita (1980$)

KenyaJordan

EthiopiaSudan

TanzaniaZaire

Nepal

PakistanBangladesh

Nigeria

MoroccoEl Salvador

Algeria

India

Indonesia

Egypt

Philippines

Sri LankaThailand

PeruTurkey

Colombia

Costa RicaMalaysia

Mexico Venezuela

Brazil

Argentina

ChileKoreaChina

Yugoslavia

Greece Hong Kong

Singapore

Trinidad and Tobago

Norm for 98 developing countries, 1982

TABLE 1TOTAL FERTILITY RATES IN THE LATE 1980S

Total Fertility RateSub-Saharan Africa 6–8India 4.2China 2.3Japan, and Western industrial democracies

1.5–1.9

Source: World Bank (1990).

6 It will be noticed that the curve in the accom-panying figure does not fit the data well. It willalso be noticed that, among the poor countries inthe sample, total fertility rates in Asia are belowwhat the curve would predict, while those in Af-rica are above. Part of my intention in this articleis to discuss possible reasons for these systematicdifferences.

1882 Journal of Economic Literature, Vol. XXXIII (December 1995)

Page 5: The Population Problem: Theory and Evidence · 2012-06-09 · The Population Problem: Theory and Evidence By PARTHA DASGUPTA University of Cambridge and Beijer International Institute

But there is a problem with it. Con-ventional indicators of the standard ofliving pertain to commodity production,not to the environmental resource-baseupon which all production ultimately de-pends. This base includes soil and itscover, freshwater, breathable air, fisher-ies, and forests. Statistics on past move-ments of gross income and agriculturalproduction say nothing about this base.They do not say if, for example, increasesin gross national product (GNP) perhead are not being realized by means ofa depletion of natural capital; in particu-lar, if increases in agricultural produc-tion are not being achieved by a “min-ing” of the soil. By concentrating onGNP (and other current-welfare mea-sures, such as life expectancy at birth),economists have, wrongly, by-passed theconcerns that ecologists have repeatedlyexpressed about the links that exist be-tween continual population growth, in-creased output, and the state of the envi-ronment. Dasgupta and Geoffrey Heal(1979), Henry Peskin (1981), and Das-gupta and Karl-Göran Mäler (1991) haveshown precisely how national product, ifit is to function effectively as an index ofsocial well-being, should include thevalue of changes in the environmentalresource-base. They have also shownthat, when it is properly defined, this in-dex, which measures net national prod-uct (NNP), takes into account the effectof changes in stocks of natural capital onfuture consumption possibilities. Thus itis possible for GNP per head to increasefor an extended period even while NNPper head is declining. We should be in aposition to say if this has been happeningin poor countries. But the practice of na-tional-income accounting has lagged sofar behind its theory, that we have littleidea of what the facts have been.7

Conventional statistics can also lull usinto thinking that human ingenuity canbe relied upon to overcome the stressesthat growing populations would other-wise impose on the natural environment;that, for every vanishing resource therewould always be substitutes for our use.But analyses of biodiversity tell us whythis would be an error. Ecologists stressthat biodiversity plays a central role inthe evolution of ecosystems (Wilson1992). Not only does biodiversity providethe units through which both energy andmaterials flow (thus giving the system itsfunctional properties, such as those ofpreserving and regenerating soil, recy-cling nutrients, pollinating crops, and fil-tering pollutants), it also provides anecosystem with resilience.

Resilience is the capacity of an ecosys-tem to recover from perturbations,shocks, and surprises. If a system losesits resilience, it can flip to a wholly newstate when subjected to even a small per-turbation. (This is where threshold ef-fects assume importance.) Thus, fieldstudies suggest that an ecosystem that isdiverse in its biota withstands stress(e.g., occasioned by violent events), inthat it manages to sustain much of itsfunctions even when the composition ofspecies changes. (Resilience should,therefore, be thought of as functional, asopposed to structural, stability of com-plex systems.) This is because there arespecies that are “waiting in the wings” totake over the functions of those that aredenuded or destroyed. Now this is remi-niscent of the assumption of substitut-

7 Attempts at estimating NNP, thus defined, arenow being made at the World Bank. However, theprinciples underlying the construction of NNP

that I have sketched in the text assume that pro-duction possibility sets are convex. This is a badassumption: there would appear to be significantnonlinearities in ecological processes (e.g., thresh-old effects; see Edward Wilson 1992; Arrow et al.1995). This means that NNP on its own is not anadequate index of social well-being. We need tosupplement NNP with estimates of the stocks ofappropriate environmental resources. On this seeDasgupta (1993, ch. 10).

Dasgupta: The Population Problem 1883

Page 6: The Population Problem: Theory and Evidence · 2012-06-09 · The Population Problem: Theory and Evidence By PARTHA DASGUPTA University of Cambridge and Beijer International Institute

ability among inputs in commodity pro-duction, an assumption that is oftenmade in economic models of technologi-cal processes. But resilience presupposesthat there are species waiting in thewings. So, to invoke the idea of substitut-ability among natural resources in com-modity production in order to play downthe utilitarian value of biodiversity, as isfrequently done (e.g., Julian Simon 1981,1994), is to make a mistake. The ac-counting price of natural capital couldwell be a good deal in excess of what it isimplicitly taken to be. For example, Pe-ter Vitousek et al. (1986) have estimatedthat 40 percent of the net energy createdby terrestrial photosynthesis (i.e., the netprimary production of the biosphere) iscurrently being appropriated for humanuse. To be sure, this is a very rough esti-mate; moreover, net terrestrial primaryproduction is not exogenously given andfixed. But the figure does put the scaleof the human presence on the planet inperspective.

In stressing the importance of biodi-versity, popular writings on the environ-ment and population growth have takena global, future-oriented view: they haveemphasized the deleterious effects that alarge population would have on ourplanet in the future. This slant, as wehave just seen, has been instructive, butit has drawn attention away from theeconomic misery endemic in large partsof the world today. Disaster is not some-thing the poorest have to wait for; it isoccuring even now. Moreover, in poorcountries, decisions on fertility and onallocations concerning education, food,work, health-care, and on the use of localenvironmental resources are in largemeasure reached and implementedwithin households that are unencum-bered by compulsory schooling, visitsfrom social workers, and so forth. Thenew perspective—the subject of this arti-cle—studies the interface of population

growth, poverty, and environmental deg-radation from a myriad of household,and ultimately individual, viewpoints. Itadopts a micro-cosmic, contemporaryoutlook, rather than a macroscopic, fu-turistic one.

III. The Household, Power, and Gender

The household as a concept is notwithout its difficulties. It is often takento mean a housekeeping or consumptionunit. The household in this sense is theeating of meals together by members, orthe sharing of meals derived from a com-mon stock of food (John Hajnal 1982).This definition has the merit that it is inaccordance with most modern censuses,but there is a problem with it: in ruralcommunities it does not yield exclusiveunits.8 A household shares a “table” andmay, for example, include live-in ser-vants who do not cook for themselves. Inmany cases some meals are had in com-mon, while others are not; and often rawand cooked food is passed to parents inadjacent cottages, apartments, and evenrooms. The boundaries vary with con-text, especially where food is not con-sumed together round a table (as inEurope) but in bowls in distinct groups(as in sub-Saharan Africa). In none ofthese cases is the housekeeping unit thesame as the consumption unit, nor is theconsumption unit necessarily well-de-fined.

In his famous analysis of fertility dif-ferences between seventeenth and eight-eenth century Northwest Europe on theone hand and modern pre-industrial so-cieties on the other, Hajnal (1982) drewupon the distinction between “nuclear”and “joint” household systems. He ob-served that in Northwest Europe mar-riage normally meant establishing a newhousehold, which implied that the cou-

8 Here I am borrowing from Goody (1990,1991).

1884 Journal of Economic Literature, Vol. XXXIII (December 1995)

Page 7: The Population Problem: Theory and Evidence · 2012-06-09 · The Population Problem: Theory and Evidence By PARTHA DASGUPTA University of Cambridge and Beijer International Institute

ple had to have, by saving or transfer,sufficient resources to establish andequip the new household. This require-ment in turn led to late marriages. It alsomeant that parents bore the cost of rear-ing their children. Indeed, fertility ratesin France dropped before mortality ratesregistered a decline, before modern fam-ily planning techniques became avail-able, and before women became literate(Ansley Coale 1969). Hajnal contrastedthis with the Asiatic pattern of householdformation, which he saw as joint unitsconsisting of more than one couple andtheir children.

Whether this two-way classification isultimately fruitful for interpreting differ-ences in fertility behavior between thetwo regions in question is open to doubt:it does not easily come to grips withmodern fertility declines in South, East,and Southeast Asia.9 Across several con-tinents the matter is even more doubtful.For example, it can be argued that therules of inheritance are a critical factorgoverning interpersonal relations, andthat differences in inheritance rules ex-plain in large measure why households insub-Saharan Africa are strikingly dissimi-lar to the “joint” household system thathas for long been taken to be the hall-mark of the Asiatic form. But inheritancerules themselves require explanation,and it is tempting to search for this inthe mode and technology of agriculturalproduction (viz. hoe versus the plough),and thereby in ecological factors (e.g.,soil quality, population density, rainfall,and availability of domestic animals).These are delicate matters of historicalanalysis, and the causal links are notwell-understood.10 Fortunately, for our

purposes here, we can by-pass such ana-lytical problems by studying instead theunderlying motivations of those agencieswhose decisions most affect fertility.Rather than regard the household as theunit of analysis, it proves more fruitful tostart with individuals, view them asagencies of action, and study theirchoices in the context of the personaland social ties to which they are, or theybecome, attached. Seen from this per-spective, the relative costs and benefitsof procreation to the various agenciesdiffer much across societies, in ways thatwe will study in this article.

Traditionally, economists have takenhouseholds to mean decision-units con-cerned with the allocation of consump-tion, work, leisure, health-care, educa-tion, and fertility. In so doing, they havedeveloped an idealized version of theconcept to explore how choices made bythese units would respond to changes inthe outside world, such as employmentopportunities, credit, insurance, health-care, and education facilities (Becker1981; Inderjit Singh, Lyn Squire, andJohn Strauss 1986). The view is ideal-ized, in that the process by which deci-sions are reached in the householdmodel is taken to be a benign one. Thus,it has been customary to interpret house-hold behavior by assuming that itschoices reflect a unitary view among itsmembers of what constitutes their well-being: the model is that of a utility maxi-mizing household. This is as true of gen-eral competitive analysis (e.g., GerardDebreu 1959) as it is of the new eco-nomic demography (e.g., Marc Nerlove,Assaf Razin, and Efraim Zadka 1987).However, if this, now traditional, in-terpretation were correct, householdchoices would be independent of which

9 On the Asian experience in recent years, seeRichard Leete and Iqbal Alam (1993).

10 See, in particular, Goody (1976, 1990) for asearching investigation. Prabhu Pingali and HansBinswanger (1987) have also offered an economicrationale for hoe agriculture in sub-Saharan Africain terms of population density and soil quality. See

Mark Rosenzweig, Binswanger, and John McIntire(1988) for further analyses of the effect of popula-tion density on agrarian relations more generally.

Dasgupta: The Population Problem 1885

Page 8: The Population Problem: Theory and Evidence · 2012-06-09 · The Population Problem: Theory and Evidence By PARTHA DASGUPTA University of Cambridge and Beijer International Institute

member actually does the choosing. Un-fortunately, the assumption is at variancewith recent findings, which have re-vealed, for example, that income in thehands of the mother has a bigger effecton her family’s health (e.g., nutritionalstatus of children) than income underthe control of the father (Eileen Ken-nedy and Ruth Oniang’o 1990). More-over, in some parts of the world (e.g.,the northern parts of the Indian subcon-tinent), household choices have beenfound to reflect allocations that favorsome members (that is, these inequali-ties cannot be explained by differencesin needs); for example, men and boysover women and girls, and the youngover the elderly.11 Household decisionswould assume strong normative signifi-cance only if the underlying basis uponwhich they are made took each mem-ber’s interests into reasonable account.But this may not be common; at least notwhen the family is impoverished and thestresses and strains of hunger, illness,and physical weakness make themselvesfelt. Thus, there is a case for movingfrom unitary to collective models of thehousehold (see also Dasgupta 1993;Harold Alderman et al. 1995).

Because we know that gender inequi-ties often prevail in education, food, andhealth-care allocations, it should not besurprising that they prevail over fertilitychoices as well. Here also men typicallywield greater influence, even thoughwomen bear the greater cost. To grasphow great the burden can be, considerthat in sub-Saharan Africa the total fer-tility rate is between six and eight (Table1). Now each successful birth involves atleast a year and a half of pregnancy andbreast-feeding. On making the obviouscorrections, we can then conclude that in

a society where female life expectancy atbirth is 50 years, and where the total fer-tility rate is, say, seven, about half of awoman’s adult life is spent either carry-ing a child in her womb or breast-feed-ing it. And we have not allowed for un-successful pregnancies.

Another indicator of the price thatwomen pay is maternal mortality. Inmost poor countries, complications re-lated to pregnancy constitute the largestsingle cause of death among women intheir reproductive years. In sub-SaharanAfrica (e.g., Ethiopia), maternal mortal-ity rates as high as one in 50 have beenrecorded.12 We may conclude that, at atotal fertility rate of seven and over, thechance that a woman who enters her re-productive years will die because of com-plications related to pregnancies is aboutone in six. Producing children thereforeinvolves playing a kind of Russian rou-lette. This is one manifestation of thepopulation problem.

As the cost they incur for procreationis so high, one expects that, given achoice, women would opt for fewer chil-dren. Thus birth rates should be ex-pected to be lower in societies wherewomen are more empowered. Data onthe status of women from 79 so-calledThird World countries (Table 2) confirmthis and display an unmistakable pat-tern: high fertility, high rates of femaleilliteracy, low share of paid employ-ment, and a high percentage working athome for no pay—they all hang together.From the data alone, it is of course diffi-cult to discern which of these measuresare causing, and which are merely corre-lated with, high fertility. But the findingsare consistent with the possibility thatlack of paid employment and educationlimits a woman’s ability to make deci-

11 Lincoln C. Chen, Emdadul Huq, and StanD’Souza (1981) is a pioneering quantitative study.Dasgupta (1993) contains references to what isnow an extensive literature on these matters.

12 By way of contrast, we should note that thematernal mortality rate in Scandinavia today is oneper 20,000.

1886 Journal of Economic Literature, Vol. XXXIII (December 1995)

Page 9: The Population Problem: Theory and Evidence · 2012-06-09 · The Population Problem: Theory and Evidence By PARTHA DASGUPTA University of Cambridge and Beijer International Institute

sions and therefore promotes populationgrowth.

The beneficial effects of parents’ edu-cation, particularly mothers’ education,on the well-being of their children havebeen much documented.13 Studies sug-gest also that education helps mothers toprocess information more effectively,and enables them to use the various so-cial and community services that may beavailable more intensively. Among otherthings, education appears to impart a de-gree of self-confidence to a person, en-abling her to avail herself of whatevernew facilities that may be on offer. Thisis invaluable for rural populations livingthrough changing circumstances.14

The links between female education,

especially secondary education, and re-productive behavior are, however, varied(Susan Cochrane 1983). The acquisitionof education delays the age of marriage,which would be expected to reduce fer-tility. Moreover, at low levels of educa-tion and contraceptive prevalence, liter-acy and receptiveness to new ideaswould be expected to complement theefforts of family planning programs.15

Furthermore, family planning programshave been known to result in longerbirth-spacing. This in turn results in a re-duction in infant mortality rates. Turningto a different set of links, education in-creases women’s opportunities for work,and thus their opportunity cost of time:the cost of child-rearing is higher foreducated mothers. And finally, educatedmothers would be expected to value edu-cation for their children more highly,and so would be more likely to make aconscious tradeoff between the qualityand number of their children. Each ofthese forces would be expected to re-duce fertility rates.

Set against these is an effect on fertil-ity that runs the other way. Taboosagainst postpartum female sexual activ-ity, where they exist, may well be weak-ened through education. In sub-SaharanAfrica, where polygyny is widely prac-ticed, postpartum female sexual absti-nence can last up to three years afterbirth. It is also not uncommon forwomen to practice total abstinence oncethey have become grandmothers. Theevidence is curious: in Latin Americaand Asia, increased female enrollment inprimary school has had the effect of low-ering fertility rates, while in many partsof sub-Saharan Africa there is evidencethat the effect has been the opposite. Ta-ble 2 displays this curiosity.

Since 1960 total fertility rates in a

TABLE 2FERTILITY RATES AND WOMEN’S STATUS

N TFR PE UE I

9 >7.0 10.6 46.9 65.735 6.1–7.0 16.5 31.7 76.910 5.1–6.0 24.5 27.1 46.025 <5.0 30.3 18.1 22.6

Source: IIED/WRI (1987, Table 2.3).Key: N: number of countries TFR: total fertility rate PE: women’s share of paid employment (%) UE: percentage of women working as unpaid family workers I: women’s illiteracy rate (%)

13 They have investigated household consump-tion of nutrients, birth-spacing, the use of contra-ceptives, child health in general, infant- and child-survival rates, and children’s height (see Dasgupta1993, ch. 12, for references). However, not all thestudies are methodologically immune to criticism.Indeed, in a few studies endogenous variables aretreated as though they are exogenous. Strauss(1990) has a good discussion of such failings.

14 Here is an indication of orders of magnitude.The infant mortality rate in households in Thai-land where the mother has had no education(resp. has had primary and secondary education)was found to be 122 per 1000 (resp. 39 and 19 per1000). See World Bank (1991).

15 Above low levels, however, female educationand family-planning outreach activities appear tobe substitutes.

Dasgupta: The Population Problem 1887

Page 10: The Population Problem: Theory and Evidence · 2012-06-09 · The Population Problem: Theory and Evidence By PARTHA DASGUPTA University of Cambridge and Beijer International Institute

number of poor countries have declinedby 50 percent or more. Schultz (1993)has shown that family planning outreachactivities explain little of either cross-country variations in fertility or changesover time in fertility within a country.He has found that the level and gendercomposition of education (affecting, forexample, child mortality), the extent ofagricultural employment (affecting thecost of raising children), and the level ofnutrition (affecting, once again, childmortality), taken together, explain mostof both types of variation.

As noted earlier, there are reasons forthinking that the extent of economic de-pendency of women on men also plays arole in fertility decisions. This depen-dency is enormous in the Indian subcon-tinent, especially in the north (David So-pher 1980; Tim Dyson and Mick Moore1983). In patrilineal societies womenrightly perceive sons as having especiallyhigh value as insurance against personalcalamities, such as widowhood and aban-donment. But sons cannot be guaran-teed. So one has to keep trying. In Eastand Southeast Asia, South India, and SriLanka women’s economic dependencywould appear to be less. Among theworld’s poorest countries in the early1970s, fertility rates have fallen mostdramatically there.16 In a wide-rangingessay on the old-age-security hypothesis(see below), Mead Cain (1984) used themedian age difference between spousesas an index of female economic depen-dence in patrilineal societies to demon-strate a remarkably high correlation be-tween this and the total fertility rate in across-sectional study of nations. Thus, inpoor societies marked by gender differ-

ences in employment opportunities andpower, women’s professed reproductivegoals need not necessarily differ notice-ably from those of men despite the dif-ferences in reproductive costs incurredby men and women (Karen Mason andAnju Taj 1987).

Opportunities for paid employmentamong women would appear to empowerthem more than educational opportuni-ties (Mayra Buvinic 1994). This has im-plications for policy. It is all well andgood, for example, to urge governmentsin poor countries to invest in literacyprograms. But the results could be disap-pointing. Many factors militate againstpoor households taking advantage of sub-sidized education. If children are neededfor work inside and outside the home,for example, then keeping them inschool (even a subsidized one) is costlyfor poor households. In places that areexperiencing high unemployment ratesamong school leavers, the benefits arealso uncertain. In patriarchal societies,educated girls can be perceived as lesspliable and harder to marry off. Weshould expect a gender-bias in educa-tional attainment. Today it is a common-place that there is such a bias (see e.g.,the World Tables in World Bank 1991).Indeed, the benefits of subsidies even toprimary education are captured dispro-portionately by families that are better off.

In contrast, policies aimed at increas-ing women’s productivity at home andimproving their earnings in the market-place would directly empower women,especially within the household. Greaterearning power would also raise for menthe implicit costs of procreation (whichkeeps women from bringing in cash in-come). One would then expect familiesto have fewer children. This is not todeny the value of public investment inprimary and secondary education inpoor countries. It is only to say that weshould be wary of claims that such in-

16 Perhaps the most impressive among these re-gions is the state of Kerala in India (where cur-rently the total fertility rate is an astonishing low,1.8). For a quantitative analysis of the determi-nants of Kerala’s demographic success, see P. N.Mari Bhat and S. Irudaya Rajan (1990).

1888 Journal of Economic Literature, Vol. XXXIII (December 1995)

Page 11: The Population Problem: Theory and Evidence · 2012-06-09 · The Population Problem: Theory and Evidence By PARTHA DASGUPTA University of Cambridge and Beijer International Institute

vestment is a panacea for the populationproblem.

The nature of income-earning oppor-tunities influences the extent to whichwomen are able to exercise choice, be-cause neither the habit of taking deci-sions nor the confidence with which oneis able to make decisions comes easily:they are both acquired. Employment as amanual worker does not involve decisionmaking in quite the extent as does self-employment as an entrepreneur. Seenfrom this perspective, employment pro-grams involving loans for commercialventures (as, say, those developed by theGrameen Bank in Bangladesh) are moreuseful than public-works.17

The importance of gender inequalityto overpopulation in poor nations is for-tunately gaining international recogni-tion. Indeed, the focal point of theUnited Nations Conference on Popula-tion and Development in Cairo in Sep-tember 1994, namely, women’s repro-ductive rights and the means by whichthey could be protected and promoted, isconsonant with the new perspective. Butthe Cairo Conference came very near totreating the problems as identical. Thiswas a mistake. There is more to thepopulation problem than gender-in-equalities, a fact that has been uncov-ered by inquiring into the various instru-ments that are available to people forcontrolling fertility and into the motivesthey have for procreation. These mattersare discussed next.

IV. Birth Control

Except under conditions of extremenutritional stress, nutritional status doesnot appear to affect fecundity (John Bon-gaarts 1980). During the 1974 famine in

Bangladesh the rural population lostover 1.5 million additional children. Thestock was replenished within a year(Bongaarts and Cain 1981). Of course,undernourishment can still have an ef-fect on sexual reproduction, through itsimplications for the frequency of still-births, maternal and infant mortality,and a possible reduction in the fre-quency of sexual intercourse.

A most obvious determinant of fertilityis the available technology for birth con-trol. Cross-country regressions (e.g.,Lant Pritchett 1994) confirm that thefraction of women of reproductive agewho use modern contraceptives isstrongly and negatively correlated withtotal fertility rates.18 So it should not besurprising that family planning programsare often seen as a prerequisite for anypopulation policy. But these regressionresults mean only that contraceptionis a proximate determinant of fertility,not a causal determinant. They couldmean, for example, that differences infertility rates across nations reflect dif-ferences in fertility goals, and therebydifferences in contraceptive use. Never-theless, it is frequently held that highfertility is a consequence of an unmetneed for modern birth-control devices,as in the claim “contraceptives arethe best contraceptive” (Bryant Robey,Shea Rutstein, and Leo Morris 1993,p. 35).

To social scientists, this is altogethertoo mono-causal an account for comfort.It is also at variance with recent findings(such as those of Schultz, 1993, men-tioned earlier) which suggest that, evenin poor countries, it is parental demand

17 The case of sub-Saharan Africa (especiallyWest Africa) is, as in many other matters, differ-ent. Women do most of the (subsistence) agricul-tural work there, but have little control over theirreproduction (see Section V).

18 For example, in East Asia over 65 per cent ofmarried women in the age range 15–49 years usecontraceptives as against somewhat under ten per-cent in sub-Saharan Africa. In South Asia as awhole the figure in the early 1980s was about 25percent, but in Sri Lanka it was a high 55 percent.Sri Lanka has one of the lowest fertility rates inAsia (see World Bank 1984, Table 7.1).

Dasgupta: The Population Problem 1889

Page 12: The Population Problem: Theory and Evidence · 2012-06-09 · The Population Problem: Theory and Evidence By PARTHA DASGUPTA University of Cambridge and Beijer International Institute

for children rather than an unmet needfor contraceptives that in large measureexplains reproductive behavior. We willinquire into the various reasons why pa-rental demand would be expected to behigh in those regions where fertility ratesare high and low where the rates are low.But it is useful to bear in mind that peo-ple in all societies practice some form ofbirth control: fertility is below the maxi-mum possible in all societies. Even inpoor countries, fertility is not unrespon-sive to the relative prices of goods andservices. As we noted earlier, extendedbreast-feeding and postpartum femalesexual abstinence have been commonpractice in Africa. In a study on !KungSan foragers in the Kalahari region,Richard Lee (1972) observed that thenomadic, bush-dwelling women amongthem had an average birth-spacing ofnearly four years, while those settled atcattle-posts gave birth to children atmuch shorter intervals. From the view-point of the individual nomadic !KungSan woman, it is significant that the so-cial custom is for mothers to nurse theirchildren on demand, and to carry themduring their day-long trips in search ofwild food through the children’s fourthyear of life. Anything less than a four-year birth interval would, therefore, in-crease mothers’ carrying loads enor-mously, impose a threat on their owncapacity to survive, and reduce theirchildren’s prospects of survival. In con-trast to bush dwellers, cattle-post womenare sedentary, and are able to wean theirchildren earlier.

Traditional methods of birth controlhave included abortion, abstinence orrhythm, coitus interruptus, and pro-longed breast-feeding.19 These options

are often inhumane and unreliable: mod-ern contraceptives are superior. Never-theless, successful family planning pro-grams have proved more difficult toinstitute than could have been thoughtpossible at first. In a wide-ranging com-mentary on the findings of the WorldFertility Survey, Cochrane and SamirFarid (1989) observed that in parts ofsub-Saharan Africa modern contracep-tive methods have been used by house-holds for altering the spacing of births,not so much the number of births. Theauthors remark that the high levels offertility in sub-Saharan Africa are a con-sequence of early and universal mar-riage, allied to little reliance on contra-ception. As Table 3 shows, the averageage at marriage for sub-Saharan womenis 18.9, and the proportion of those 15–19 years of age who are married is 40percent. The corresponding figures forAsia are 21.3 and 26 percent, respec-tively, and for Latin America, 21.5 yearsand 19.8 percent, respectively. But theproximate determinants identified byCochrane and Farid are themselves inneed of explanation, and we will lookinto it. Sub-Saharan Africa’s populationgrowth rate has increased during thetwo decades 1960–80 from about 2.5percent per year to something like 2.9percent per year. We will see below thatthe absence in Africa of a strong conju-gal bond (in particular the practice of

TABLE 3

A (years) B (%)

sub-Saharan Africa 18.9 40 Asia 21.3 26 Latin America 21.5 19.8

Source: Cochrane and Farid (1989).Key A: average age of marriage for women B: proportion of women aged 15–19 years who are married

19 Anthropologists have, however, argued thatin sub-Saharan Africa prolonged breast-feeding isnot a birth-control measure, but a means of reduc-ing infant mortality: traditionally, animal milk hasbeen scarce in the region.

1890 Journal of Economic Literature, Vol. XXXIII (December 1995)

Page 13: The Population Problem: Theory and Evidence · 2012-06-09 · The Population Problem: Theory and Evidence By PARTHA DASGUPTA University of Cambridge and Beijer International Institute

polygyny20) has something to do withsuch high rates, as does the concept ofthe self and its ties with kinship (Goody1976; Meyer Fortes 1978). But it is notclear if any of this has much to do withthe fact that, barring a few countries,fertility rates there have not shown anysignificant decline in recent years, de-spite a decline in infant mortality rates.The importance of women in farming hasoften been adduced to explain, in part,sub-Saharan Africa’s marriage patterns.But, as we observed in the previous sec-tion, differences between the genders inthe net benefits of having children are akey ingredient in the population problemfacing both the Indian subcontinent andsub-Saharan Africa, not just the latter.

Households would be expected toadopt new methods of birth control tosatisfy unmet needs. However, over timeit is the net demand for children thatwould be expected to dominate house-hold decisions. We should not be sur-prised, therefore, that, in those regionswhere family planning programs havehad an impact, it has occurred mostly inthe initial stages. Here is a substantia-tion: Starting in 1977, 70 “treatment” vil-lages were serviced by a program of birthcontrol in the famous study in MatlabThana in Bangladesh, while 79 “control”villages were offered no such special ser-vice. The contraceptive prevalence in thetreatment villages increased from sevento 33 percent within 18 months, andthen more gradually to a level of 45 per-cent by 1985. The prevalence also in-creased in the control villages, but onlyto 16 percent in 1985. By 1980 the dif-ference in total fertility rates betweenthe two groups reached a figure of 1.5(Kenneth Hill 1992). The question iswhether the Matlab experiment shouldbe viewed as an exception, or whether it

could be replicated easily in otherplaces.

In a notable paper, Pritchett (1994)has argued that it would be unrealistic touse the Matlab program as an indicatorof the effectiveness of birth-control pro-grams, for the reason that it was bothmassive and expensive. The fertility re-sponse was certainly large, but he esti-mates that the cost of each preventedbirth was about 120 percent of Bangla-desh’s GDP per capita, a very high fig-ure.

Pritchett (1994) also analyzed datafrom household surveys conducted bythe World Fertility Survey and theDemographic and Health Surveys pro-grams, which included women’s re-sponses to questions regarding both theirpreferences and their behavior on fertil-ity matters. Demographers had earlierderived several indicators of the demandfor children from these data. One suchindicator, the “wanted total fertility rate”(Bongaarts 1990), can be compared tothe actual total fertility rate for the pur-pose of classifying births or current preg-nancies in a country or region as wantedor unwanted. Regressing actual fertilityon fertility desires on a sample of 43countries in Asia, Africa, and LatinAmerica, Pritchett found that about 90percent of cross-country differences intotal fertility rates are associated withdifferences in desired fertility. More-over, excess fertility was found not to besystematically related to the actual totalfertility rate, nor to be an important de-terminant of the rate. The figure 90 per-cent may prove to be an over-estimate,but it is unlikely to prove to be greatlyso.21 Even in poor households, the use ofmodern contraceptives would involveonly a small fraction (one percent orthereabouts) of income.

20 In West Africa between 40 and 50 percent ofwives are in polygynous marriages (John and PatCaldwell 1990).

21 I am grateful to John Bongaarts for helpfulconversations on this matter.

Dasgupta: The Population Problem 1891

Page 14: The Population Problem: Theory and Evidence · 2012-06-09 · The Population Problem: Theory and Evidence By PARTHA DASGUPTA University of Cambridge and Beijer International Institute

It is reasonable to conclude, then, thateven in poor countries fertility rates to-day are dependent mostly (perhaps 70–80 percent) on the net demand for chil-dren and not on the extent of familyplanning outreach activities. The impor-tance international agencies continue toplace on family planning programs as away of eliciting lower fertility rates is atodds with these findings. We should in-stead be seeking to understand the de-terminants of the demand for children.This is taken up next.

V. Children as Ends

One motive for procreation, commonto humankind, relates to children as endsin themselves. We are genetically pro-grammed to want and to value them. Ithas also been said that children are theclearest avenue open to “self-transcen-dence” (David Heyd 1992).22 Viewingchildren as ends ranges from the desireto have children because they are playfuland enjoyable, to a desire to obey thedictates of tradition and religion. Onesuch injunction emanates from the cultof the ancestor, which, taking religion tobe the act of reproducing the lineage, re-quires women to bear many children.23

This latter motivation has been empha-sized by Caldwell and Caldwell (1990) to

explain why sub-Saharan Africa has forthe most part proved so resistent to fer-tility reduction.

The problem with this argument isthat, although it explains why fertilityrates in sub-Saharan Africa are high, itdoes not explain why the rates have notresponded to declines in mortality. Thecult of the ancestor may prescribe repro-duction of the lineage, but it does notstipulate an invariant fertility rate. Evenin sub-Saharan Africa, total fertility rateshave been below the maximum possiblerate; so they should be expected to re-spond to declines in mortality, a matter Ishall come back to in Section VI, and inSection VII I will offer one possible ex-planation for the resistance that sub-Sa-haran Africa has shown to reductions infertility rates.24

The view that children are an end inthemselves provides several pathways bywhich reasoned fertility decisions at thelevel of every household could lead to anunsatisfactory outcome from the per-spectives of all households.25 One suchpathway arises from the fact that tradi-tional practice is often perpetuated by

22 Note that in evolutionary biology phenotypiccosts and benefits of reproduction are importantonly to the extent that they are correlated withreproductive measures. Offspring in this theoryare valued in terms of the end of increasing fit-ness. This is not the point of view in economicdemography.

23 Writing about West Africa, Fortes (1978, pp.125–26) says “a person does not feel he has ful-filled his destiny until he or she not only becomesa parent but has grandchildren . . . [Parenthood] isalso a fulfillment of fundamental kinship, religiousand political obligations, and represents a commit-ment by parents to transmit the cultural heritageof the community . . . Ancestry, as juridicallyrather than biologically defined, is the primary cri-terion . . . for the allocation of economic, political,and religious status.” See also Goody (1976). Co-chrane and Farid (1989) remark that both the ur-

ban and rural, the educated and uneducated insub-Saharan Africa have more, and want more,children than their counterparts do in other re-gions. Thus, even the younger women there ex-pressed a desire for an average of 2.6 more chil-dren than women in the Middle East, 2.8 morethan women in North Africa, and 3.6 to 3.7 morethan women in Latin America and Asia.

24 Between 1965 and 1987 the infant mortalityrate in a number of the poorest countries in sub-Saharan Africa declined from about 200 per 1,000live births to something like 150 per 1,000 livebirths (World Bank 1989). Caldwell, I. O. Orubu-loye, and Caldwell (1992) suggest that the declinesin fertility rates that have been observed in recentyears in Botswana, Kenya, Zimbabwe, and parts ofNigeria are due to further declines in mortalityrates that these places have recently enjoyed.

25 Reproductive externalities have not beenmuch studied in the “new economic demography”so far. Surveying the field, Schultz (1988, p. 417)writes: “Consequences of individual fertility deci-sions that bear on persons outside of the familyhave proved difficult to quantify, as in many caseswhere social external diseconomies are thought tobe important.”

1892 Journal of Economic Literature, Vol. XXXIII (December 1995)

Page 15: The Population Problem: Theory and Evidence · 2012-06-09 · The Population Problem: Theory and Evidence By PARTHA DASGUPTA University of Cambridge and Beijer International Institute

imitative behavior. Procreation in closelyknit communities is not only a privatematter, it is also a social activity, influ-enced by both family experiences andthe cultural milieu (Goody 1976, 1990;Fortes 1978; Richard Easterlin, RobertPollak, and Michael Wachter 1980;Susan Cotts Watkins 1990).

Formally speaking, imitative behaviorwould occur if every household’s mostdesired family size were an increasingfunction of the average family size in thecommunity.26 This is, of course, a “re-duced form” of the concept (Dasgupta1993, ch. *12), and the source of a desireto imitate could lie in reasons other thanan intrinsic desire to be like others. Itcould be that similar choices made byhouseholds generate mutual positive ex-ternalities, say, because people careabout their status, and a household’schoice of actions signals its predisposi-tions, and thereby affects its status (B.Douglas Bernheim 1994).

Whatever the basis of imitative behav-ior, there would be practices encourag-ing high fertility rates that no householdwould unilaterally desire to break. Suchpractice could well have had a rationalein the past, when mortality rates werehigh, rural population densities werelow, the threat of extermination fromoutside attack was large, and mobilitywas restricted. But practices can surviveeven when their original purpose has dis-appeared. It can then be that, so long asall others follow the practice and aim atlarge family sizes, no household on itsown wishes to deviate from the practice;however, if all other households were torestrict their fertility rates, each woulddesire to restrict its fertility rate as well.In other words, imitative behavior can bea reason for the existence of multipleequilibria.

This said, it must also be acknowl-edged that testing for multiple equilibriais a most difficult matter. For the mo-ment it is analytical reasoning that tellsus that a society could in principle getstuck at a self-sustaining mode of behav-ior, characterized by high fertility andlow educational attainment, even whenthere is another, potentially self-sustain-ing, mode of behavior that is charac-terized by low fertility and high educa-tional attainment (Dasgupta 1993, ch.*12).

This does not mean that society wouldbe stuck with high fertility rates forever.As always, people differ in the extent oftheir absorption of traditional practice.There would inevitably be those who, forone reason or another, experiment, takerisks, and refrain from joining the crowd.They are the tradition-breakers, and theyoften lead the way. Educated women areamong the first to make the move towardsmaller families (see e.g., Ghazi Farooq,Ita Ekanem, and Sina Ojelade, 1987, fora commentary on West Africa). Femaleeducation is thus a potent force for cre-ating tradition-breakers, as are earningopportunities for women. A concertedsocial effort toward furthering theseends could help dislodge such a societyfrom the rapacious hold of high fertilityrates and low educational attainment to amode of behavior where fertility is lowand educational attainment high. A moredirect route would be social coordinationthrough education and public exhorta-tion, aimed particularly at men if the so-ciety is strongly patriarchal. In this con-text, the role of television and radio intransmitting information about other lifestyles can be especially important(Ronald Freedman 1995).

VI. Children as Productive Assets

Still other motives for procreation in-volve viewing children as productive as-

26 This is an instance of “strategic complemen-tarities.” See Russell Cooper and Andrew John(1988).

Dasgupta: The Population Problem 1893

Page 16: The Population Problem: Theory and Evidence · 2012-06-09 · The Population Problem: Theory and Evidence By PARTHA DASGUPTA University of Cambridge and Beijer International Institute

sets. In a rural economy where the ave-nues for saving are highly restricted, orwhere public support for the elderly areweak, parents value children as a sourceof security in their old age. Cain (1981,1983) has studied this aspect extensivelyon the basis of data from Bangladesh,and Jeffrey Nugent and Thomas Gillaspy(1983) have used Mexican evidence toargue that old-age pension and social se-curity do act as a substitute for children.One way of formalizing this is to assumethat parents are interested in householdwelfare, subject to the requirement thatthe chance of there being an offspring tocare for them in old age (i.e., providingsustenance, time, and attention) is noless than a certain amount. In many re-gions (e.g., the Indian subcontinent) thistranslates itself to a requirement that thechance of there being a son alive whenthe parents are old is no less than a cer-tain amount. As a numerical example, wemay consider the simulation study byDavid May and David Heer (1968), whoestimated that an average Indian couplein the 1960s needed to have 6.3 childrenin order to be 95 percent sure of havinga surviving son when the father reachesthe age of 65. This is a high figure, aboutthe same as the total fertility rate in In-dia during the decade of the 1950s.27

Old-age security provides a potentiallystrong motive. In 1980 people aged 65and over in South Asia formed aboutfour percent of the total population. Thesex composition among the aged is farfrom even, being of the order of 80–85men for every 100 women among theelderly. In South and Southeast Asia fe-male life expectancy at birth is 59 years,while that of males is about 54 years; atage 60, however, they are approximately15 and 14 years, not much less than the

life expectancy at age 60 in advanced in-dustrial countries. In the Indian subcon-tinent the proportions of the elderly wholive with their children (for the mostpart, sons) is of the order of 80 percentor more. (In the United States the corre-sponding figure is about 15 percent.)Sons are a necessity in these circum-stances. A poor widow with no sons innorthern parts of the Indian subconti-nent is faced with the prospect of desti-tution.28

Related to this is a phenomenon thathas been observed by Jane Guyer (1994)in a Yoruba area of Nigeria. In the faceof deteriorating economic circumstances,some women are bearing children by dif-ferent men so as to create immediate lat-eral links with them. Polyandrous moth-erhood enables women to have access tomore than one resource network.

Children differ in their potential. Onewould expect parents in a poor house-hold to develop the most promising oftheir children, even if this were to meanthat the remaining ones are somewhatmarginalized. This is confirmed by botheconomic theory and evidence (Beckerand Nigel Tomes 1976; Caroline Bledsoe1994). Daughters are a net drain on pa-rental resources in the Indian subconti-nent (dowries can be bankrupting). Thisgoes some way toward explaining thepreference parents show for sons there(Sopher 1980; Cain 1984). It also helpsexplain why daughters receive less edu-cation and are expected to work rela-tively harder for their parents, a matterwe come to next.

27 Samuel Preston (1978) is a useful collectionof essays on the effects that have been observedon fertility rates of reductions in rates of infantmortality.

28 Urbanization tends to break households downinto “nuclear” units, thereby raising the parentalcosts of procreation (Coale and Edgar Hoover1958; Schultz 1993). Urbanization in a growingeconomy also offers children better employmentprospects, which improve their bargainingstrength relative to their parents. This in turn low-ers the return on children as investment, becausechildren become less dependable as a source ofincome to their parents in their old age.

1894 Journal of Economic Literature, Vol. XXXIII (December 1995)

Page 17: The Population Problem: Theory and Evidence · 2012-06-09 · The Population Problem: Theory and Evidence By PARTHA DASGUPTA University of Cambridge and Beijer International Institute

In poor countries children are alsouseful as income-earning assets. Thisprovides households in these parts withanother motive for procreation. It hasconsequences that have only recentlybeen explored in theoretical analyses(Dasgupta and Mäler 1991, 1995; Ner-love 1991; Nerlove and Anke Meyer1993).

Poor countries in great part arebiomass-based subsistence economies.Rural folk there eke out a living fromproducts obtained directly from plantsand animals. Production throughput islow. Much labor is needed even for sim-ple tasks. Moreover, households there donot have access to the sources of domes-tic energy available to households in ad-vanced industrial countries. Nor do theyhave water on tap. In semi-arid and aridregions, water supply is often not evenclose at hand. Nor is fuel wood near athand when the forests recede. Thismeans that the relative prices of alterna-tive sources of energy and water faced byrural households in poor countries arequite different from those faced byhouseholds elsewhere. In addition to cul-tivating crops, caring for livestock, cook-ing food, and producing simple market-able products, members of a householdmay have to spend as much as five to sixhours a day fetching water and collectingfodder and wood. These are complemen-tary activities. They have to be under-taken on a daily basis if the household isto survive. Each is time-consuming. La-bor productivity is low not only becausecapital is scarce, but also because envi-ronmental resources are scarce. Fromabout the age of six years, children inpoor households in poor countries mindtheir siblings and domestic animals,fetch water, and collect fuel wood, dung,and fodder. Children are then needed asworkers by their parents, even when theparents are in their prime. In their studyof work allocation among rural house-

holds in the foothills of the Himalayas,the Centre for Science and Environment(C.S.E 1990) recorded that children be-tween 10 and 15 years work one-and-a-half times the number of hours adultmales do, their tasks consisting of col-lecting fuel wood, dung, and fodder,grazing domestic animals, performinghousehold chores, and marketing.29 In-deed, children can add so much tohousehold income that, in some places,they are costless to rear by the time theyreach adolescence. Cain (1977) studieddata from the village Char Gopalpur inBangladesh. He estimated that male chil-dren become net producers at as early anage as 12 years, and work as many hoursa day as an adult. Using a zero (calorie)rate of interest, he calculated that malechildren compensate for their own cu-mulative consumption by the age of 15.This may not be typical in Bangladesh. Icite it, nevertheless, to show the vast dif-ference in the motivation for having chil-dren between households in rich coun-tries and poor households in poorcountries. Each household would appearto need many hands, and it can be thatthe overall usefulness of each additionalhand increases with declining resourceavailability, at least over some range.30

VII. Positive Feedbacks and PovertyTraps

The need for many hands can lead to adestructive situation, especially whenparents do not have to pay the full priceof rearing their children but share thosecosts with their community. In recentyears, mores that once regulated local re-sources have changed. Since time imme-morial, rural assets such as village pondsand water holes, threshing grounds, graz-

29 See Julia Falconer and J. E. A. Arnold (1989)for a similar investigation for sub-Saharan Africa.

30 This can happen if households discount thefuture at a sufficiently high rate. See Nerlove(1991).

Dasgupta: The Population Problem 1895

Page 18: The Population Problem: Theory and Evidence · 2012-06-09 · The Population Problem: Theory and Evidence By PARTHA DASGUPTA University of Cambridge and Beijer International Institute

ing fields, and local forests have beenowned communally. This form of owner-ship and control enabled households insemi-arid regions to pool their risks.James Howe (1986), Robert Wade(1988), Kanchan Chopra, Gopal Kadek-odi, and M. N. Murty (1990), Jean-MarieBaland and Jean-Philippe Platteau(1995), and others have shown that, tra-ditionally, communities have protectedlocal commons from overexploitation byrelying on norms, imposing fines for de-viant behavior, and other means. But thevery process of economic developmentcan erode traditional methods of control,say, by way of increased urbanization andmobility (Dasgupta 1993, ch. 10). Socialnorms are also endangered by civil strifeand by the usurpation of resources bylandowners or the state. As norms de-grade, parents pass some of the costs ofchildren on to the community by over-exploiting the commons. Indeed, even amarginal decline in compliance in agree-ments can trigger a process of cumula-tive causation. Over time the effectcould be large. If access to shared re-sources continues, parents produce toomany children, which leads to greatercrowding and susceptibility to disease aswell as to more pressure on local envi-ronmental resources. But no household,on its own, would take into account theharm it would inflict on others whenbringing forth another child. This is an-other instance of a demographic free-rider problem.

Parental costs of procreation are alsolower when the cost of rearing the childis shared among the kinship. In sub-Sa-haran Africa, “fosterage” within the kin-ship is a commonplace: children are notraised solely by their parents, the re-sponsibility is more diffuse within thekinship group (Goody 1976; Bledsoe1990). Fosterage in the African contextis not adoption. It is not intended to, nordoes it in fact, break ties between par-

ents and children. The institution affordsa form of mutual insurance protection insemi-arid regions. There is some evi-dence that, as savings opportunities arefew in the low-productivity agriculturalregions of sub-Saharan Africa, fosteragealso enables households to smoothentheir consumption across time.31 In partsof West Africa up to half the childrenhave been found to be living with their kinat any given time. Nephews and nieceshave the same rights of accommodationand support as do biological offspring.There is a sense in which children areseen as common-responsibility. However,the arrangement creates yet anotherfree-rider problem if the parents’ shareof the benefits from having children ex-ceeds their share of the costs. From thepoint of view of the parents, taken as acollective, too many children would beproduced in these circumstances.32

31 This latter motivation has been explored byRenata Serra in a graduate thesis at the Universityof Cambridge.

32 To see that there is no distortion if the shareswere the same, suppose c is the cost of rearing achild and N the number of couples within a kin-ship. For simplicity assume that each child makesavailable y units of output (this is the norm) to theentire kinship, which is then shared equally amongall couples, say in their old age. Suppose also thatthe cost of rearing each child is shared equally byall couples. Let n* be the number of children eachcouple other than the one under study chooses tohave. (We will presently endogenize this.) If nwere to be the number of children this couple pro-duces, it would incur the resource cost C = [nc +(N − 1)n*c]/N, and eventually the couple wouldreceive an income from the next generation equal-ling Y = [ny + (N − 1)n*y]/N. Denote the couple’saggregate utility function by the form U(Y) −K(C), where both U(.) and K(.) are increasing andstrictly concave functions. Letting n be a continu-ous variable for simplicity, it is easy to confirmthat the couple in question will choose the valueof n at which yU′ (Y) = cK′ (C). The choice sus-tains a social equilibrium when n = n*. (This is thesymmetric non-cooperative Nash equilibrium ofthe social system.) It is easy to check that this isalso the condition which is met in a society wherethere is no reproductive free-riding. It is a simplematter to confirm that there is free-riding if theparents’ share of the benefits from having childrenexceeds their share of the costs.

1896 Journal of Economic Literature, Vol. XXXIII (December 1995)

Page 19: The Population Problem: Theory and Evidence · 2012-06-09 · The Population Problem: Theory and Evidence By PARTHA DASGUPTA University of Cambridge and Beijer International Institute

In sub-Saharan Africa, communal landtenure of the lineage social structure of-fers yet another inducement for men toprocreate. In addition, as conjugal bondsare weak, fathers often do not bear thecosts of siring a child. Anthropologistshave observed that the unit of Africansociety is a woman and her children,rather than parents and their children.Often, there is no common budget forthe man and woman. Descent in sub-Sa-haran Africa is, for the most part, patri-lineal and residence is patrilocal (an ex-ception are the Akan people of Ghana).Patrilineality, weak conjugal bonds, com-munal land tenure, and a strong kinshipsupport system of children, taken to-gether, are a broad characteristic of theregion. In principle they provide a pow-erful stimulus to fertility. Admittedly,patrilineality and patrilocality are fea-tures of the northern parts of the Indiansubcontinent also. But conjugal bondsare substantially greater there. More-over, as agricultural land is not commun-ally held, large family sizes lead to frag-mentation of landholdings. In contrast,large families in sub-Saharan Africa are(or, at least were, until recently) re-warded by a greater share of land be-longing to the lineage or clan.

The perception of both low costs andhigh benefits of procreation in sub-Saha-ran Africa induces “couples” to producetoo many children. Theoretical consid-erations suggest that, in certain circum-stances, a disastrous process can therebybegin. As the community’s natural re-sources are depleted, more hands areneeded to gather fuel and water for dailyuse. More children are then produced,further damaging the local resource baseand in turn providing the “household”with an incentive to enlarge. When thishappens, poverty, fertility, and environ-mental degradation reinforce one an-other in an escalating spiral. By the timesome countervailing set of factors—

whether public policy or diminishedbenefits from having further childrendue, say, to a scarcity of land—stops thespiral, millions of lives may have sufferedthrough worsening poverty.33

Kevin Cleaver and Götz Schreiber(1994) provide some evidence for thisthesis in the context of rural sub-SaharanAfrica. They report positive correlationsbetween poverty, fertility, and deteriora-tion of the local environmental resourcebase. Such data cannot reveal causal con-nections, but they are not inconsistentwith the idea of a positive-feedbackmechanism such as I have described.Over time, this spiral would be expectedto have large effects, as manifested bybattles for resources (William Durham1979; Thomas Homer-Dixon, JeffreyBoutwell, and George Rathjens 1993).

And there is evidence that is more in-direct. The victims of such a process aswe are discussing would be born andraised in poverty. A large proportionwould suffer from undernourishment.They would remain illiterate and wouldoften be both stunted and wasted. Un-dernourishment would retard their cog-nitive (and often motor) developmentand compromise their future capacity towork. Labor productivity would be dis-mally low. For the most part investmentcredit would be unavailable to them sothat savings would be severely con-strained. Stunting and wasting wouldhave a tendency of begetting stuntingand wasting down the generations of adynasty (Dasgupta 1993, chs. 14–16).Among those who would survive, the vic-tims hit hardest would be society’s out-casts—the migrants and dispossessed,some of whom in time would become theemaciated beggars seen on the streets oflarge towns and cities in poor countries.Nutritional findings (John Waterlow

33 Nerlove (1991) has provided a formal analysisof such positive feedback processes.

Dasgupta: The Population Problem 1897

Page 20: The Population Problem: Theory and Evidence · 2012-06-09 · The Population Problem: Theory and Evidence By PARTHA DASGUPTA University of Cambridge and Beijer International Institute

1986, 1992), historical studies (RobertFogel 1994a, 1994b) and theoretical ex-plorations (Dasgupta 1993, ch. 16;1995b), when taken together, show thatthe spiral I have outlined here is one wayin which destitutes are created. Emaci-ated beggars are not lazy; they have tohusband their precarious hold on energy.Having suffered from malnutrition, theycease to be marketable.

Families with greater access to re-sources would, however, be in a positionto limit their size and propel themselvesinto still higher income levels. I have notbeen able to locate published data on thematter, but my impression is that amongthe urban middle classes in north Indiathe transition to a low fertility rate hasalready been achieved. This does notmean there is an inexorable “vicious cir-cle of poverty.” People from the poorestof backgrounds have been known to liftthemselves out of the mire. Neverthe-less, there are forces at work which pullhouseholds away from one another interms of their living standards. India pro-vides a possible example of how the vi-cious cycle I have described can enableextreme poverty to persist amid a growthin well-being in the rest of society. TheMatthew Effect (“For unto everyone thathath shall be given, and he shall haveabundance; but from him that hath notshall be taken away even that which hehath”) would appear to work relentlesslyin poor countries.

In this background, it is hard to makesense of the oft-expressed suggestion(e.g., Simon 1981) that there are cumula-tive benefits to be enjoyed from in-creases in population size even in poorcountries; that human beings are a valu-able resource. To be sure, they are po-tentially valuable as doers of things andoriginators of ideas, but for this they re-quire the means for personal develop-ment. Moreover, historical evidence onthe way pressure of population led to

changes in the organization of produc-tion, property rights, and ways of doingthings, which is what Ester Boserup(1981) studied in her far-reaching work,also does not seem to speak to the popu-lation problem as it exists today in sub-Saharan Africa and the northern parts ofthe Indian subcontinent.

VIII. Public Policy

The analysis presented here suggeststhat the way to reduce fertility would beto break the destructive spiral wheresuch a spiral is in operation. Because pa-rental demand for children, rather thanan unmet need for contraceptives, ingreat measure explains reproductive be-havior in poor countries, we should try toidentify policies that would so changethe options men and women face thattheir reasoned choice would be to lowertheir fertility.

In this regard, civil liberties, as op-posed to coercion, would appear to playa particular role. In Dasgupta (1990) Ishowed, by the use of statistical analysisof data pertaining to the decade of the1970s from 51 of the then poorest coun-tries, that political and civil liberties arepositively and significantly correlatedwith improvements in income per head,life expectancy at birth, and the infantsurvival rate. Correlation is not causa-tion, but there are now reasons for think-ing that such liberties are not only desir-able in themselves, but also haveinstrumental virtues in empowering peo-ple to flourish in the economic sphere.The causal chain may well be that politi-cal and civil liberties provide sustenanceto the rule of law, and thereby to secu-rity of property and the enforcement ofcontracts. In fact Adam Przeworski andFernando Limongi (1995) have shownthat fertility, as well, is lower in coun-tries where citizens enjoy more civil andpolitical liberties. (An exception is

1898 Journal of Economic Literature, Vol. XXXIII (December 1995)

Page 21: The Population Problem: Theory and Evidence · 2012-06-09 · The Population Problem: Theory and Evidence By PARTHA DASGUPTA University of Cambridge and Beijer International Institute

China, which represents only one out ofsome 100 countries in their sample.)

The most potent solution in semi-aridregions of sub-Saharan Africa and the In-dian subcontinent is to deploy a numberof policies simultaneously. Family plan-ning services, especially when allied withhealth services, and measures that em-power women are certainly desirable. Associal norms break down and traditionalsupport systems falter, those women whochoose to change their behavior becomefinancially and socially more vulnerable.So a literacy and employment drive forwomen is essential to smooth the transi-tion to lower fertility. But improving so-cial coordination and directly increasingthe economic security of the poor arealso essential. Providing infrastructuralgoods, such as cheap fuel and potablewater, will reduce the usefulness of extrahands. When a child becomes perceivedas expensive, we may finally have a hopeof dislodging the rapacious hold of highfertility rates. Neither evidence noranalysis has yet disproved the notion thatthe poor in poor countries know, at leastin a rough manner, what is in their self-interest. But each of the prescriptionsoffered by our new perspective is desir-able by itself, and not just when we havethe population problem in mind. Itseems to me that this consonance ofmeans and ends is a most agreeable factin what is otherwise a depressing field ofstudy.

Admittedly, in saying all this we arelooking at matters wholly from the per-spective of the parents. This is limiting.34

But developing the welfare economics ofpopulation policies has proved to be ex-

tremely difficult (Dasgupta 1994): ourethical intuition at best extends to actualand future people, we do not yet possessa good moral vocabulary for includingpotential people in the calculus. What Ihave tried to argue in this essay is thatthere is much that we can establish evenif we were to leave aside such conceptualdifficulties. Population policy involves agood deal more than making family plan-ning centers available to the rural poor.It also involves more than a recognitionthat poverty is a root cause of high fertil-ity rates. The problem is deeper, but as Ihave tried to show, it is possible to sub-ject it to analysis.

REFERENCES

ALDERMAN, HAROLD ET AL. “Unitary versus Col-lective Models of the Household: Is it Time toShift the Burden of Proof?” World Bank Re-search Observer, Feb. 1995, 10(1), pp. 1–20.

ARROW, KENNETH ET AL. “Economic Growth,Carrying Capacity, and the Environment,” Sci-ence, 28 Apr. 1995, 268(5210), pp. 520–21.

BALAND, JEAN-MARIE AND PLATTEAU, JEAN-PHILIPPE. Halting degradation of natural re-sources: Is there a role for rural communities?Oxford: Oxford U. Press, 1995.

BECKER, GARY. A treatise on the family. Cam-bridge, MA: Harvard U. Press, 1981.

BECKER, GARY S. AND TOMES, NIGEL. “ChildEndowments and the Quantity and Quality ofChildren,” J. Polit. Econ., Supplement, Aug.1976, 84(4, Part 2), pp. S143–62.

BERNHEIM, B. DOUGLAS. “A Theory of Conform-ity,” J. Polit. Econ., Oct. 1994, 102(5), pp. 841–77.

BHAT, P. N. MARI AND RAJAN, S. IRUDAYA.“Demographic Transition in Kerala Revisited,”Economic and Political Weekly, 1–8 Sept. 1990,25(35&36), pp. 1957–80.

BIRDSALL, NANCY. “Economic Approaches toPopulation Growth,” in Handbook of develop-ment economics. Vol. 1. Eds.: HOLLIS CHEN-ERY AND T. N. SRINIVASAN. Amsterdam: NorthHolland, 1988, pp. 477–542.

BIRDSALL, NANCY AND JAMISON, DEAN T. “In-come and Other Factors Influencing Fertility inChina,” Population Dev. Rev., Dec. 1983, 9(4),pp. 651–75.

BLEDSOE, CAROLINE. “The Politics of Children:Fosterage and the Social Management of Fertil-ity Among the Mende of Sierra Leone,” inBirths and power: Social change and the poli-tics of reproduction. By W. PENN HAND-WERKER.London: Westview Press, 1990, pp.81–89.

34 Stephen Enke (1966) is a notable explorationof the value of prevented births when the worth ofadditional lives is judged to be based entirely ontheir effect on the current generation. As a simpli-fication, Enke took the value of a prevented birthto be the discounted sum of the differences be-tween an additional person’s consumption andoutput over the person’s lifetime.

Dasgupta: The Population Problem 1899

Page 22: The Population Problem: Theory and Evidence · 2012-06-09 · The Population Problem: Theory and Evidence By PARTHA DASGUPTA University of Cambridge and Beijer International Institute

———. “‘Children are Like Young BambooTrees’: Potentiality and Reproduction in sub-Saharan Africa,” in KERSTIN LINDAHL-KIES-SLING AND HANS LANDBERG 1994, pp. 105–38.

BONGAARTS, JOHN. “Does Malnutrition Affect Fe-cundity? A Summary of the Evidence,” Science,9 May 1980, 208(4444), pp. 564–69.

———. “The Measurement of Wanted Fertility,”Population Dev. Rev., Sept. 1990, 16(3), pp.487–506.

BONGAARTS, JOHN AND CAIN, MEAD. Demo-graphic responses to famine. New York: Popula-tion Council, 1981.

BOSERUP, ESTER. Population growth and techno-logical change. Chicago: Chicago U. Press,1981.

BUVINIC, MAYRA. “Population Policy and FamilyPlanning Programmes: Contribution from a Fo-cus on Women,” in Population—the complex re-ality. Ed.: FRANCIS GRAHAM-SMITH. London:The Royal Society, 1994, pp. 211–28.

CAIN, MEAD T. “The Economic Activities of Chil-dren in a Village in Bangladesh,” PopulationDev. Rev., Sept. 1977, 3(3), pp. 201–27.

———. “Risk and Insurance: Perspectives on Fer-tility and Agrarian Change in India and Bangla-desh,” Population Dev. Rev., Sept. 1981, 7(3),pp. 435–74.

———. “Fertility as an Adjustment to Risk,”Population Dev. Rev., Dec. 1983, 9(4), pp. 688–702.

———. “Women’s Status and Fertility in Develop-ing Countries: Son Preference and EconomicSecurity.” World Bank Staff Working Paper No.682, World Bank, Washington, DC, 1984.

CALDWELL, JOHN C. AND CALDWELL, PAT.“High Fertility in sub-Saharan Africa,” Scien-tific American, May 1990, 262(5), pp. 82–89.

CALDWELL, JOHN C.; ORUBULOYE, I. O. ANDCALDWELL, PAT. “Fertility Decline in Africa: ANew Type of Transition?” Population Dev. Rev.,June 1992, 18(2), pp. 211–42.

CHEN, LINCOLN C.; HUQ, EMDADUL ANDD’SOUZA, STAN. “Sex Bias in the Family Alloca-tion of Food and Health Care in Rural Bangla-desh,” Population Dev. Rev., Mar. 1981, 7(1),pp. 55–70.

CHOPRA, KANCHAN; KADEKODI, GOPAL ANDMURTY, M. N. Participatory development: Peo-ple and common property resources. NewDelhi: Sage Publications, 1990.

CLEAVER, KEVIN M. AND SCHREIBER, GÖTZ A.Reversing the spiral: The population, agricul-ture, and environment nexus in sub-Saharan Af-rica. Washington, DC: World Bank, 1994.

COALE, ANSLEY J. “The Decline of Fertility inEurope from the French Revolution to WorldWar II,” in Fertility and family planning: Aworld view. Eds.: S. J. BEHRMAN, LESLIECORSA, JR., AND RONALD FREEDMAN. Ann Ar-bor, Michigan: U. of Michigan Press, 1969, pp.3–24.

COALE, ANSLEY J. AND HOOVER, EDGAR M.Population growth and economic development

in low-income countries. Princeton, NJ: Prince-ton U. Press, 1958.

COCHRANE, SUSAN. “Effects of Education and Ur-banization on Fertility,” in Determinants of fer-tility in developing countries. Vol II. Eds.RUDOLFO BULATAO AND RONALD LEE. NewYork: Academic Press, 1983, pp. 587–626.

COCHRANE, SUSAN AND FARID, SAMIR. “Fertilityin Sub-Saharan Africa: Analysis and Explana-tion.” World Bank Discussion Paper No. 43,Washington DC, 1989.

COOPER, RUSSELL AND JOHN, ANDREW. “Co-ordinating Coordination Failure in KeynesianModels,” Quart. J. Econ., Aug. 1988, 103(3),pp. 441–63.

C.S.E. (Centre for Science and Environment). Hu-man-nature interactions in a Central Himala-yan Village: A case study of village Bemru. NewDelhi: Centre for Science and Environment,1990.

DASGUPTA, PARTHA. “Well-Being and the Extentof its Realisation in Poor Countries,” Econ. J.,Mar. 1990, 100(400, Supplement), pp. 1–32.

———. An inquiry into well-being and destitu-tion. Oxford: Clarendon Press, 1993.

———. “Savings and Fertility: Ethical Issues,”Philosophy & Public Affairs, Spring 1994, 23(2),pp. 99–127.

———. “Population, Poverty and the Local Envi-ronment,” Scientific American, Feb. 1995a,272(2), pp. 40–45.

———. “Nutritional Status, the Capacity forWork, and Poverty Traps.” Mimeo. Faculty ofEconomics, U. of Cambridge, 1995b; forthcom-ing, J. Econometrics.

DASGUPTA, PARTHA AND HEAL, GEOFFREY. Eco-nomic theory and exhaustible resources. Cam-bridge: Cambridge U. Press, 1979.

DASGUPTA, PARTHA AND MÄLER, KARL-GÖRAN.“The Environment and Emerging DevelopmentIssues,” Proceedings of the annual bank confer-ence on development economics 1990 (Supple-ment to the World Bank Economic Review),1991, pp. 101–32.

———. “Poverty, Institutions, and the Environ-mental-Resource Base,” in Handbook of devel-opment economics. Vol. III. Eds.: JEREBEHRMAN AND T. N. SRINIVASAN. Amsterdam:North Holland, 1995, pp. 2371–2463.

DEBREU, GERARD. Theory of value. New York:John Wiley, 1959.

DRÉZE, JEAN AND SEN, AMARTYA. Hunger andpublic action. Oxford: Clarendon Press, 1990.

DURHAM, WILLIAM. Scarcity and survival in Cen-tral America: Ecological origins of the SoccerWar. Stanford, CA: Stanford U. Press, 1979.

DYSON, TIM AND MOORE, MICK. “On KinshipStructure, Female Autonomy, and Demo-graphic Behavior in India,” Population Dev.Rev., Mar. 1983, 9(1), pp. 35–60.

EASTERLIN, RICHARD; POLLAK, ROBERT ANDWACHTER, MICHAEL. “Toward a More GeneralModel of Fertility Determination: EndogenousPreferences and Natural Fertility,” in Popula-

1900 Journal of Economic Literature, Vol. XXXIII (December 1995)

Page 23: The Population Problem: Theory and Evidence · 2012-06-09 · The Population Problem: Theory and Evidence By PARTHA DASGUPTA University of Cambridge and Beijer International Institute

tion and economic change in developing coun-tries. Ed.: RICHARD EASTERLIN. Chicago: U. ofChicago Press, 1980, pp. 81–135, 144–49.

EHRLICH, PAUL R. AND EHRLICH, ANNE H. Thepopulation explosion. New York: Simon &Schuster, 1990.

ENKE, STEPHEN. “The Economic Aspects of Slow-ing Population Growth,” Econ. J., Mar. 1966,76(1), pp. 44–56.

FALCONER, JULIA AND ARNOLD, J. E. M. House-hold food security and forestry: An analysis ofsocio-economic issues. Rome: Food and Agricul-ture Organization, 1989.

FAROOQ, GHAZI; EKANEM, ITA I. AND OJELADE,SINA. “Family Size Preferences and Fertility inSouth-Western Nigeria,” in Sex roles, popula-tion and development in West Africa. Eds.:CHRISTINE OPPONG. London: James Currey,1987, pp. 75–85.

FOGEL, ROBERT. “Economic Growth, PopulationTheory, and Physiology: The Bearing of Long-Term Processes on the Making of EconomicPolicy,” Amer. Econ. Rev., June 1994a, 84(3),pp. 369–95.

———. “The Relevance of Malthus for the Studyof Mortality Today: Long-Run Influences onHealth, Mortality, Labour Force Participation,and Population Growth,” 1994b in KERSTENLINDAHL-KIESSLING AND HANS LANDBERG1994, pp. 231–84.

FORTES, MEYER. “Parenthood, Marriage and Fer-tility in West Africa,” J. Devel. Stud., July 1978,14(4, Special Issue on Population and Develop-ment), pp. 121–49.

FREEDMAN, RONALD. “Asia’s Recent Fertility De-cline and Prospects for Future DemographicChange.” Asia-Pacific Population Research Re-port No. 1, East-West Center, Honolulu, 1995.

GOODY, JACK. Production and Reproduction.Cambridge: Cambridge U. Press, 1976.

———. The oriental, the ancient, and the primi-tive. Cambridge: Cambridge U. Press, 1990.

———. “Comparison of Family and Demographyin Europe and Asia: thoughts on the Hajnal hy-pothesi.” Mimeo. St. John’s College, Cam-bridge, 1991.

GUYER, JANE I. “Lineal Identities and LateralNetworks: The Logic of Polyandrous Mother-hood,” in Nupitality in sub-Saharan Africa:Contemporary anthropological and demo-graphic perspectives. Eds.: CAROLINE BLED-SOE AND GILLES PISON. Oxford: ClarendonPress, 1994, pp. 231–52.

HAJNAL, JOHN. “Two Kinds of PreindustrialHousehold Formation Systems,” PopulationDev. Rev., Sept. 1982, 8(3), pp. 449–94.

HEYD, DAVID. Genethics: The morality of pro-creation. Los Angeles: U. of Calfornia Press,1992.

HILL, KENNETH. “Fertility and Mortality Trendsin the Developing World,” Ambio, Feb. 1992,21(1), pp. 79–83.

HOMER-DIXON, THOMAS; BOUTWELL, JEFFREYAND RATHJENS, GEORGE. “Environmental

Change and Violent Conflict,” Scientific Ameri-can, Feb. 1993, 268(2), pp. 16–23.

HOWE, JAMES. The Kuna gathering: Contempo-rary village politics in Panama. Austin, Texas:U. of Texas Press, 1986.

IIED/WRI (International Institute for Environ-ment and Development/World Resources Insti-tute). World resources 1987. New York: BasicBooks, 1987.

KELLEY, ALLEN C. “Economic Consequences ofPopulation Change in the Third World,” J.Econ. Lit., Dec. 1988, 26(4), pp. 1685–728.

KENNEDY, EILEEN AND ONIANG’O, RUTH.“Health and Nutrition Effects of SugarcaneProduction in South-Western Kenya,” Food andNutrition Bulletin, Dec. 1990, 12(4), pp. 261–67.

LEE, RICHARD. “Population Growth and the Be-ginnings of Sedentary Life among the !KungBushmen,” in Population growth: Anthropologi-cal implications. Ed.: BRIAN SPOONER. Cam-bridge, MA: MIT Press, 1972, pp. 329–42.

LEETE, RICHARD AND ALAM, IQBAL, eds. Therevolution in Asian fertility: Dimensions, causesand implications. Oxford: Clarendon Press,1993.

LINDAHL-KIESSLING, KERSTIN AND LANDBERG,HANS, eds. Population, economic development,and the environment: The making of our com-mon future. Oxford and New York: Oxford U.Press, 1994.

MASON, KAREN AND TAJ, ANJU. “Differences be-tween Women’s and Men’s Reproductive Goalsin Developing Countries,” Population Dev.Rev., Dec. 1987, 13(4), pp. 611–38.

MAY, DAVID A. AND HEER, DAVID M. “Son Survi-vorship Motivation and Family Size in India: AComputer Simulation,” Population Studies, July1968, 22(2), pp. 199–210.

NATIONAL ACADEMY OF SCIENCES. Rapid popu-lation growth: Consequences and policy impli-cations. Baltimore: Johns Hopkins U. Press,1971.

NATIONAL RESEARCH COUNCIL. Populationgrowth and economic development: Policy ques-tions. Washington, DC: National Academy ofSciences Press, 1986.

NERLOVE, MARC. “Population and the Environ-ment: A Parable of Firewood and Other Tales,”Amer. J. Agr. Econ., Dec. 1991, 73(4), pp.1334–47.

NERLOVE, MARC AND MEYER, ANKE. “En-dogenous Fertility and the Environment: A Par-able of Firewood,” in “The Environment andEmerging Development Issues.” Eds.: PARTHADASGUPTA AND KARL-GÖREN MÄLER.Mimeo. Faculty of Economics, U. of Cam-bridge, 1993; forthcoming, Oxford: ClarendonPress.

NERLOVE, MARC; RAZIN, ASSAF AND SADKA,EFRAIM. Household and economy: Welfare eco-nomics of endogenous fertility. New York: Aca-demic Press, 1987.

NUGENT, JEFFREY B. AND GILLASPY, R. THOMAS.

Dasgupta: The Population Problem 1901

Page 24: The Population Problem: Theory and Evidence · 2012-06-09 · The Population Problem: Theory and Evidence By PARTHA DASGUPTA University of Cambridge and Beijer International Institute

“Old Age Pension and Fertility in Rural Areasof Less Developed Countries: Some Evidencefrom Mexico,” Econ. Devel. Cult. Change, July1983, 31(4), pp. 809–29.

PESKIN, HENRY. “National Income Accounts andthe Environment,” Natural Res. J., July 1981,21(3), pp. 511–37.

PINGALI, PRABHU AND BINSWANGER, HANS.“Population Density and Agricultural Intensifi-cation: A Study of the Evolution of Technolo-gies in Tropical Agriculture,” in Populationgrowth and economic development: Issues andevidence. Eds.: D. GALE JOHNSON ANDRONALD LEE. Madison, WI: U. of WisconsonPress, 1987, pp. 27–56.

PRESTON, SAMUEL H., ed. The effects of infantand child mortality on fertility. New York: Aca-demic Press, 1978.

PRITCHETT, LANT H. “Desired Fertility and theImpact of Population Policies,” Population Dev.Rev., Mar. 1994, 20(1), pp. 1–56.

PRZEWORSKI, ADAM AND LIMONGI, FERNANDO.“Democracy and Development.” Working Paper#7, Chicago Center on Democracy, U. of Chi-cago, 1995.

ROBEY, BRYANT; RUTSTEIN, SHEA O. AND MOR-RIS, LEO. “The Fertility Decline in DevelopingCountries,” Scientific American, Dec. 1993,269(6), pp. 30–37.

ROSENZWEIG, MARK; BINSWANGER, HANS ANDMCINTIRE, JOHN. “From Land Abundance toLand Scarcity: The Effects of PopulationGrowth on Production Relations in AgrarianEconomies,” in Population, food and rural de-velopment. By RONALD LEE ET AL. Oxford:Clarendon Press, 1988, pp. 77–100.

SCHULTZ, T. PAUL. “Economic Demography andDevelopment,” in The state of development eco-nomics: Progress and perspectives. Eds.: GUS-TAV RANIS AND T. PAUL SCHULTZ. Oxford:Basil Blackwell, 1988, pp. 416–51.

———. “Sources of Fertility Decline in ModernEconomic Growth and Fertility: Is AggregateEvidence on the Demographic Transition Cred-

ible?” Mimeo. Department of Economics, YaleU., 1993.

SIMON, JULIAN L. The ultimate resource. Prince-ton: Princeton U. Press, 1981.

———. “Debate Statement,” in Scarcity or abun-dance? A debate on the environment. By NOR-MAN MYERS AND JULIAN L. SIMON. New York:W.W. Norton, 1994, pp. 5–68.

SINGH, INDERJIT; SQUIRE, LYN AND STRAUSS,JOHN, eds. Agricultural household models: Ex-tensions, applications, and policy. Baltimore:Johns Hopkins U. Press, 1986.

SOPHER, DAVID E. “Sex Disparity in Indian Liter-acy,” in An exploration of India: Geographicalperspectives on society and culture. Ed. DAVIDE. SOPHER. Ithaca, NY: Cornell U. Press, 1980,pp. 130–90.

STRAUSS, JOHN. “Households, Communities, andPreschool Children’s Nutrition Outcomes: Evi-dence from Rural Côte d’Ivoire,” Econ. Devel.Cult. Change, Jan. 1990, 38(2), pp. 231–61.

VITOUSEK, PETER ET AL. “Human Appropriationof the Product of Photosynthesis,” BioScience,June 1986, 36(6), pp. 368–73.

WADE, ROBERT. Village republics: Economic con-ditions for collective action in South India.Cambridge: Cambridge U. Press, 1988.

WATERLOW, JOHN. “Metabolic Adaptation to LowIntakes of Energy and Protein,” Annual Re-views of Nutrition, 1986, 6, pp. 495–526.

———. Protein-energy malnutrition. Sevenoaks,Kent: Edward Arnold, 1992.

WATKINS, SUSAN COTTS. “From Local to NationalCommunities: The Transformation of Demo-graphic Regions in Western Europe 1870–1960,” Population Dev. Rev., June 1990, 16(2),pp. 241–72.

WILSON, EDWARD O.The diversity of life. Cam-bridge, MA: Harvard U. Press, 1992.

WORLD BANK. World development report. NewYork: Oxford U. Press, 1984, 1990, 1991.

———. Sub-Saharan Africa: From crisis to sus-tainable growth. Washington, DC: World Bank,1989.

1902 Journal of Economic Literature, Vol. XXXIII (December 1995)