SWP692 Recent Fertility Declines in Brazil, Colombia, and Mexico Thomas W. Merrick WORLD BANK STAFF WORKING PAPERS Number 692 POPULATION AND DEVELOPMENT SERIES Number 17 FLE COPY n vi y~~~~~~~~~~~~~~~~~~~~~ 1- t;-' WORLD BANK STAFF WORKING PAPERS 3"9/. 3 Number 692 &LS7 POPULATION AND DEVELOPMENT SERIES 4 7 Number 17 JY7O- 69.X eo 49 Recent Fertility Declines in Brazil, Colombia, and Mexico Thomas W. Merrick .45~~4 The World Bank Washington, D.C., U.S.A. Copyright ©) 1985 The Intemational Bank for Reconstruction and Development/THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing January 1985 This is a working document published informally by the World Bank. To present the results of research with the least possible delay, the typescript has not been prepared in accordance with the procedures appropriate to fonnal printed texts, and the World Bank accepts no responsibility for errors. The publication is supplied at a token charge to defray part of the cost of manufacture and distribution. The World Bank does not accept responsibility for the views expressed herein, which are those of the authors and should not be attributed to the World Bank or to its affiliated organizations. The findings, interpretations, and conclusions are the results of research supported by the Bank; they do not necessarily represent official policy of the Bank. The designations employed, the presentation of material, and any maps used in this document are solely for the convenience of the reader and do not imply the expression of any opinion whatsoever on the part of the World Bank or its affiliates concerning the legal status of any country, territory, city, area, or of its authorities, or concerning the delimitation of its boundaries, or national affiliation. The full range of World Bank publications, both free and for sale, is described in the Catalog of Publications; the continuing research program is outlined in Abstracts of Current Studies. Both booklets are updated annually; the most recent edition of each is available without charge from the Publications Sales Unit, Department T, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or from the European Office of the Bank, 66 avenue d'1ena, 75116 Paris, France. Thomas W. Merrick is president of the Population Reference Bureau (Washington, D.C.) and a consultant to the World Bank. Librairy of Congrress Calaloging in 1?mbHcarion Data Merrick, Thomas William. Recent fertility declines in Brazil, Colombia, and Mexico. (World Bank staff working papers ; no. 692. Population and development series ; no, 17) Bibliography: p. 1. Fertility, Human--Brazil. 2. Fertility, Human-- Colombia.. 3. Fertility, Human--Mexico. 4. Birth control --Brazil. 5. Birth control--Colombia. 6. Birth control-- Mexico. I. Title. II. Series: World Bank staff working papers ; no. 692. III. Series: World Bank staff working papers. Population and development series ; no. 17, HB973.M46 1985 304.6'32 84-29142 ISBN 0-8213-0490-9 FOREWORD This paper is one in a special series of World Bank Staff Working Papers on population change and development. Prepared as background papers for the World Development Report 1984, they provide more detailed treatment and documentation of the issues dealt with in Part II of the Report. The papers cover a range of topics, including the effects of population growth and change on economic development, the determinants of fertility and mortality, the links between population growth and internal and international migration, and the management, financing, and effectiveness of family planning programs. They include several country and regional studies of fertility change and population policy. The background papers draw on a large number of published and unpublished studies of individual researchers, on Bank policy analysis and research, and on reports of other organizations working on population and development programs and issues. The papers are the work of individuaLs and the views and interpretations expressed in them do not necessarily coincide with the views and interpretations of the Report itself. I hope these detailed studies will supplement the World Development Report 1984 in furthering understanding of population and development issues among students and practitioners of development. Nancy Birdsall Staff Director World Development Report 1984 Some of the Papers in the Population and Development Ser-°es Ainsworth, Martha. Family Planning Programs: The Clients' Perspectlve0 World Bank Staff Working Paper no. 676. Boulier, Bryan L. Evaluating Unmet Need for Contraception: Estimates for Thirty- six Developing Countries. World Bank Staff Working Paper no. 678. Bulatao, Rodolfo A. Expenditures on Population Programs in Developing Regions: Current Levels and Future Requirements0 World Bank Staff Working Paper no. 679. Bulatao, Rodolfo A0 Reducing Fertility in Developing Countries: A Review of Determinants and Policy Leverso World Bank Staff Working Paper ro0 6800 Bulatao, Rodolfo Ao and Anne Elwan. Fertility and Mortality Transition in Developing Countries: Patterns, Projections, and Interdependence0 Wcrld Bank Staff Working Paper no0 6810 Cain, Mead. Women's Status and Fertility in DevelopLng Countries: ton Preference and Economic Securityo World Bank Staff Working Paper no0 6820 Gwatkin, Davidson0 Mortality Reduction, Fertility Decline, and Population Growth: Toward a More Relevant Assessment of Relationships Among Them0 World Bank Staff Working Paper noo 686o Herz, Barbarao Official Development Assistance for Population Activi.ties. A Review0 World Bank Staff Working Paper no0 6880 McNicoll, Geoffrey. Consequences of Rapid Population Growth: An Overview0 World Bank Staff Working Paper no0 691. Merrick, Thomas W. Recent Fertility Declines in Brazil, Colombia, and Mexico0 World Bank Staff Working Paper no. 6920 Muscat, Robert and others. Rapid Population Growth and Human Carrying Capacity' Two Perspectives0 World Bank Staff Working Paper no0 690O Sapir, Andre. Some Aspects of Population Growth, Trade, and Fector iNobility. World Bank Staff Working Paper no0 694. Standing, Guyo Population Mobility and Productive Relations: Demographic Links and Policy Evolution0 World Bank Staff Working Paper no0 6950 Tan, Jee-Peng and Michael Haines. Schooling and Demand for ChIldren: Historical Perspectives0 World Bank Staff Working Paper no0 697. Trussell, James and Anne Ro Pebley0 The Potential Impact of Changes in Fertility on Infant, Child, and Maternal Mortality0 World 3ank Staff Working Paper no0 698. Zachariah, KoCo and Sulekha Patel. Determinants of Fertility Declize in Ind'.a= An Analysis0 World Bank Staff Working Paper noo 6990 Zachariah, K0C0 The Anomoly of the Fertility Decline in India's Kerala State, A Field Investigationo World Bank Staff Working Paper no0 700O Abstract Because of decline in fertility, population growth has been declining in Brazil and Colombia since the mid-1960s and in Mexico since the mid-1970s. The rapidity of the declines is one of the most striking similarities among the three countries. They also had similar levels of urbanization, life expectancy, and female literacy, but they differed in the way in which their population policies evolved. Colombia, with the highest initial fertility and most rapid decline, was the earliest to organize large-scale family planning. In Mexico, the onset of rapid fertility decline coincided with a major change in national policy on public support of family planning. Brazil did not have a national population program, though the use of contraceptives spread through private commercial channels as well as through privately organized family planning programs. The paper examines the determinants of the onset and spread of fertility decline in each country and concludes that both supply and demand factors contributed to their acceleration. Analysis of the proximate determinants of fertility indicates increased use of effective contraceptives as the main contributor on the supply side. Demand was important because the fertility decline was also related to increased economic pressures on lower income classes. Overall, the decline in fertility has accelerated because it has spread to these income groups. Condense La baisse de la fecondit6 a entrain6 un ralentissement de la croissance d6mographique au Bresil et en Colombie des le milieu des annees 60, et au Mexique depuis le milieu des annees 70. L'une des res- semblances les plus frappantes entre les trois pays, qui avaient par ail- leurs des niveaux similaires d'urbanisation, d'esperance de vie et d'al- phabetisation des femmes, r6side dans la rapidite de ce ralentissement. Cependant, leur politique demographique a evolue de facon fort diff6rente. La Colombie, qui avait au depart le taux de f6condite le plus elev6 et a enregistre la baisse la plus rapide, a ete la premiere a mettre en place des services de planning familial a vaste echelle. Au Mexique, la baisse rapide de la fecondite a coincid6 avec un changement majeur dans la posi- tion des pouvoirs publics touchant l'appui au planning familial. Le Br6sil n'avait pas de programme national de population, mais l'emploi de contraceptifs s'est 6tendu par les circuits commerciaux prives et grace a des programmes de planning familial priv6s. Le document examine les facteurs qui ont determin6 le declenche- ment et la g6neralisation de la baisse de la f6condit6 dans chaque pays et conclut que l'offre et la demande ont toutes deux joue un r6le dans l'ac- cel6ration du processus. Du cot6 de l'offre, l'analyse des determinants imm6diats de la fecondite fait ressortir que le recours accru a des con- traceptifs efficaces a joue un r6le cl1. La demande n'est pas a negliger non plus, puisque le recul de la fecondite est lie a l'intensification des pressions economiques sur les classes a faible revenu. Dans 1'ensemble, la baisse de la f6condite s'est acceler6e parce qu'elle s'est 6tendue a ces groupes de revenu. Extracto En raz6n de los mas bajos niveles de fecundidad, el crecimiento de la poblaci6n ha disminuido en el Brasil y Colombia desde mediados del decenio de 1960, y en M6xico desde mediados del de 1970. Una de las similitudes mais notables entre dichos paises es la rapidez de la disminuci6n registrada en cada uno de ellos. Los tres tenian ademas niveles similares de urbanizaci6n, esperanza de vida y alfabetizaci6n femenina, pero diferian en la manera en que evolucionaron sus respectivas politicas de poblaci6n. Colombia, donde habia el mAs alto nivel inicial de fecundidad y se registr6 la disminuci6n mAs rapida, fue el primero en organizar la planificaci6n familiar en gran escala. En Mexico, el comienzo de la rapida disminuci6n de la fecundidad coincidi6 con un importante cambio de la politica nacional en cuanto al apoyo del sector publico a la planificaci6n familiar. El Brasil no tenia un programa nacional de poblaci6n, pero el uso de anticonceptivos se difundi6 a traves de conductos comerciales privados, asi como de programas de planificaci6n familiar organizados privadamente. En el documento se examinan los elementos determinantes del comienzo y difusi6n del descenso de la fecundidad en cada pais y se Ilega a la conclusi6n de que los factores de la oferta y la demanda contribuyeron a su aceleraci6n. El anAlisis de las causas inmediatas de la fecundidad revela un uso acrecentado de anticonceptivos eficaces como el factor principal desde el punto de vista de la oferta. La demanda revisti6 importancia puesto que la disminuci6n de la fecundidad estuvo relacionada tambien con las mayores presiones econ6micas sobre los grupos de ingresos mas bajos. En conjunto, la disminuci6n de la fecundidad se ha acelerado porque ha llegado a estos grupos. Table of Contents Page Introduction ................... . . ............ ................ 1 Explaining Recent Fertility Declines ........... .. ....... ..... 3 Fertility Trends Since 1965 ........................................ . 7 The Proximate Determinants of Total Fertility .............. 13 Explanations ... . .... ............... ...... . 22 Conclusions .................................................. 38 References .................................................. 41 Introduction.1 Brazil, Mexico, and Colombia account for 60 percent of Latin America's total population of 390 million people, and 63 percent of its 12 million annual births. All three have experienced very rapid rates of population growth during the last three decades. Average annual rates accelerated to levels of 3 percent or more during the 1950s and 1960s, when death rates had declined but birth rates remained high. Since the mid-1960s growth rates have been falling in Brazil and Colombia as a result of declines in fertility, with Mexico following in the mid-1970s. One of the most striking similarities among the three countries is the rapidity of these fertility declines (see Table 1 for comparative demographic and socio-economic measures). By the early 1980s, Colombia's birth rate was 29 per thousand, compared to 46 in early 1960s. Brazil's was 30, as against 43 per thousand, while the corresponding figures for Mexico were 32 and 45 per thousand. All three countries have official population policies, but differ as to when the policy was adopted and in the nature of the policy. According to Nortman (1982), Colombia and Mexico have official policies to reduce their population growth rates and to support family planning for thavu purpose, while Brazil is listed as supporting family planning for other purposes. Colombia adopted its official policy in 1970, but had been very permissive toward private sector and international agency support of family planning in the years prior to establishing a national policy. Mexico adopted its policy at the end of 1973. This was an abrupt shift from its earlier pronatalist position, which had been quite restrictive about outside support of family 2 Table lo Economic, Social and Demographic Indicators: Brazil9 Colombia9 and Mexico. Brazil Colombia Mexico Per capita income 2050 1180 2090 (1980 $U.S.) Average annual growth 5.1 3.0 2.6 in GNP per capita (1960- 80) Population, mid-1980 118.8 26.7 69.8 (millions) Urban population, percent of total (1960) 46 48 51 (1981) 68 64 67 Adult literacy9 percent (1960) 61 63 65 (1980) 76 81 83 Life expectancy (1960) 57 53 57 (1981) 66 63 66 Infant mortality (1960) 118 103 9i (1981) 75 55 54 Income distribution: Share of lowest 100 9.0 9.5 40% in total household income (around 1970) Percent decline in 30.2 34.8 28.9 crude birth rate (1960-1980) Source: World Development Indicators sectior of World Bank9 World Development Report, 1983 except income distribution data9 which are taken-rcr- CGYenery et aTTIT974). 3 planning. Since then it has established a National Population Council charged with stabilizing population growth, set specific growth targets, and moved vigorously to build a network of family planning clinics. Brazil incorporated a population policy statement in a 1974 National Development Plan. It called for an assessment of the social and economic impact of demographic factors, but stopped short of admitting to the need for control of population growth or providing federal govermient support for family planning programs. Brazil has been permissive about private and international agency support for family planning. Several state goverrnments have established programs in cooperation with those agencies. In Brazil, as well as Mexico and Colombia, private commercial sales of contraceptives and private sector physicians have played an important role in making family planning available to middle and upper income class urban couples. Explaining Recent Fertility Declines. The recent rapid declines of fertility in these three countries raise two basic questions. First, why did birth rates remain so high until the middle 1960s (for Mexico, the middle 1970s), even though all three by then had achieved significant levels of urbanization and advancement on other scales of modernization such as per capita income, education, and life expectancy? Second, what social and economic changes, combined with changes in reproductive attitudes and practices, led to declines in fertility that appear to be as rapid, or even more so, than the declines in mortality that occurred a few decades earlier? Are we witnessing a Latin American variant of the European demographic transition? Or does this experience represent something new in terms of the relation between demographic and socio-economic variables? 4 There has been considerable skepticism about the applicab.ility of the European transition model to Latin American experience. Arriaga '1970) wrote: The European fertility-morta7ity pattern could not have been repeated in Latin America between 1930 and 1960 principally because za the shorter period of time allowed for change in the latter. In addit on, a decline in fertility in response to falling mortality was hampered by the particular social, cultural , educational , and economic characteristics of Latin American Countries0 Coale (1978), writing on the relation between population and development in Mexico, added: The Mexican experience must also be viewed as paradoxical For those who hold the optimistic position that countries with rapid popV,ation growth need have no concern about it, because social and economic progress automatically bring down birth rates0 . . The Mexican population in 1975 was more than 72 percent literate and more than 60 Dercent urban, the per capita income had nearly doubled in 20 years, the expectation of life had risen to about 65 years and was still increasing; nevertheless, fertility until 1975 was, if anything, higher than it was 20 years before. While most commentators agreed that the Latin American context betw^een 1940 and 1970 was too complex to allow a straight-forward application of demographic transition theory to the question of when and how fertility would decline, not all were as skeptical of its validity as Arriaga and Coale. Beaver (1975) and Oeschsli and Kirk (1975) argued that the theory did yield generalizations that were applicable to Latin America. Each examined a wide range of social and economic indicators associated with modernization. Oechsli and Kirk predicted that a number of countries in the region were on the threshold of very rapid fertility decline by 19600 Beaver pestulated that once fertility decline was underway, it would be at least as rapid as the decline in mortality. Coale himself argued that a broader view of trans1ition theory was needed. In a review of the theory (1975), he concluded that three broad generalizations relevant to developing countries could be derived from the theory. They are the conditions required for onset of a fert,i;ity decline: 5 (1) That fertility must be within the calculus of conscious individual choice (as opposed to biological limitations or social mechanisms that affect fertility); (2) That fertility reduction must be perceived as advantageous by individuals making reproductive choices; and (3) That effective means of fertility control must be available. If recent fertility declines in Brazil, Colombia, and Mexico represent a demographic transition defined in Coale's terms, then one should be able to explain the persistence of high fertility until the mid-1960s as well as the shift to rapid fertility decline after that in terms (1) of the absence of one or more of the conditions with respect to broad segments of their populations until about 1965, and (2) of changes in one or more of the conditions since then. Stating these questions in terms of Coale's conditions raises further questions about how they relate to each other: i.e., is any one of the conditions both a necessary and a sufficient condition for fertility decline? For example, motivation would appear to be a necessary condition, but not sufficient without access to the means to control. Access could also be a necessary condition, bu.t not sufficient without motivation. At least two scenarios suggest themselves. One is that there was pent-up demand in these societies for smaller families, but it was not realized because access to the means of fertility control was limited. It was realized when the means finally did become more widely available. For purposes of discussion, label the supply access condition as hypothesis A. Another possibility is that one or both of the other two conditions changed. That is, in addition to changes in supply (necessary condition), there were also changes either in the degree to which couples were thinking about smaller families and what they could do to achieve them (Coale's calculus of conscious choice) or in the relative advantages and disadvantages of smaller versus 6 larger families, or both. Label the motivation condition as hypothesis B. While birth control consciousness can be conceptualized as a separate hypothesis, it is more likely that changes in this condition twiould accompany increased access to family planning or increased motivation, or a combination of both. For this reason, we do not consider it as a separate f.ypothesis. When one examines conditions during the 1960s and changes that have occurred since at the national level, there is a strong prima facie case for hypothesis A. By 1960, these countries had achieved comparatively high levels on modernization scales that social scientists had been using to measure the "demand" side of fertility: urbanization, life expectancy, literacy and educational attairnent. What was perplexing to so many observers was the persistence of high fertility rates in the face of such high levels of apparent modernization. The paper will suggest that the persistence of high fertility is much less difficult to understand if one pays attention (1) to distribution as well as the level of achievement of modernization variables and (2) to differences in the impact of macro-level socioeconomic changes on different groups which had high fertility prior to 1965 but experienced decline after. In 1960, as well as 1980, national averages masked significant rural-urban, regional and socio-economic class differentials in fertility and other demographic models in all three countries. Fertility had begun to decline in the `;odern" groups and regions, but remained high for those who had not participated as much in the modernization process0 National level fertility declines ocCurred because lower fertility spread to lower inccme classes0 The distribution issue requires us to refine hypotheses A and B0 In hypothesis A, the question is whether there was pent-up demand for smaller families among lower socio-economic classes which limited access to family 7 plannning prevented them from realizing. In hypothesis B, two variants are possible: (BI) Was it that in addition to increased access, there was also mobility from lower to higher status groups, so while class differences remained, the weight of lower class groups in the total diminished? and/or (B2) Did economic conditions affecting the reproductive motivation of lower class groups change, so that there was more to be gained (and possibly greater awareness of that gain) by controlling fertility during the 1970s than in the 1960s? Fertility Trends Since 1965. Neither Brazil, Colombia, nor Mexico have vital statistics systems that are reliable enough to provide accurate measurement of the timing and magnitude of recent fertility declines. The National Academy of Sciences' Committee on Population and Demography has published indirect fertility estimates covering the period from the mid-1960s to late-1970s for Brazil and Colombia, and similar information is available for Mexico. Table 2 presents estimates of the total fertility rate (TFR) derived from infonration in the 1960 censuses, summary estimates from the National Academy for the mid-1960s to late 1970s, as well as estimates for around 1980 that have recently become available.2 All three countries had TFRs that were 6 or more in the early 1960s. Brazil appears to have been the first to start a decline: its TFR was already lower than 6 in 1965, though the exact date of the onset of the decline is difficult to establish because estimates for the 1960s are based on retrospective reports in census and survey data from the 1970's. These are 8 Table 2. Trends in the Total Fertility Rate. DATE BRAZIL COLOMBIA MLEXICO Around 1960 6.2 6.8 6.5 1965 5.8 6.6 6.6 1970 5.3 5.7 6.7 1975 4.7 4.8 5.7 Around 1980 4.3 3.8 4.1 Percent decline* 31.0 44.0 370 Sources: Brazil and Colombia, National Academy of Sciences, Committee on Population and Demography, Reports 12 (1982) and 21 (1983); Mexico, Potter and Alba (1983), Table 3. Mexico's 1980 figure is taken from preliminary reports of its 1982 national demographic survey. Around 1960 to around 1980. subject to a variety of errors that affect timing. By 1980, Brazil's TFR had fallen 31 percent, and there were nearly 2 fewer births per woman than in 1960. Colombia had the highest TFR among the three countries in the early 1960s. According to the National Academy of Sciences, its decline did not start until the middle of the 1960s; once underway, the decrease was even greater than in Brazil. By 1980, Colombia's TFR was 3.8, which represented 3 fewer births per woman than in 1960, The onset of fertility decline in Mexico came still later, some time during the early 1970s, since total fertility was still well over 6 in 1970. A decline of nearly 40 percent occurred between 1970 and 1980 (total fertility increased during the 1960s), indicating that the pace of decline is even more rapid than in Colombia and Brazil. In both 9 Colombia and Mexico, the onset of fertility decline appears to have anticipated by a short time the establishment of large scale organized family planning efforts, both private and government sponsored, but accelerated as those programs gained momentum. In Brazil, neither the onset of fertility decline nor its acceleration during the 1970s was associated with large scale organized efforts to provide family planning. Regional and rural-urban fertility differentials characterized all three countries prior to the onset of recent declines. In Brazil, total fertility rates in Rio and Sao Paulo were three births per woman lower than in the Northeast as early as 1950, indicating that the transition to lower fertility was already underway in the Southeast by that time (Table 3). Between 1950 and 1970, total fertility actually increased in the Northeast and Frontier regions. Declines of about 0.5 births per woman occurred in Minas Gerias/Espirito Santo and the Southern region, whose 1950 fertility rates were inter-mediate between Rio and Sao Paulo, and in Sao Paulo. After 1970, declining fertility spread to the Northeast and Frontier, with each experiencing declines of about 1.5 births per woman during the decades. In Minas Gerais/Espirito Santo and the South, there were declines of about two births per woman. In tenrs of the number of births, the declines in Rio and Sao Paulo were less, since rates had been lower to begin with, though the percentage decline was substantially greater in Rio than in Sao Paulo. Thus regional differentials have narrowed during the 1970s. Earlier declines in urban total fertility rates reflect the experiences of Rio and Sao Paulo, where much of Brazil's urban population was concentrated. By 1980, 68 percent of Brazil's population counted as urban. Urban total fertility was 4.7 births per woman in 1950, and declined only slightly between 1950 and 1970. Between 1950 and 1970 rural rates remained 10 steady at 7.7 births per woman and then dropped by about 1.3 birQ'es during the last decade. Aggregate fertility levels actuaity declined more t'hFin either the urban or rural average because of the redistribution of populat on frcm rurao to urban areas. During the 1970s9 rural population in Brazil deciined in both percentage and absolute terms. Table 3, Total Fertility Rates, Regional and Rural.Urban Differentials in Brazil, 1950-80. Region 1950 1970 1976 1980 80/70 Rio de Janeiro 4,4 4.3 3.9 2D9 o67 Sao Paulo 405 4.0 3.2 3.2 .80 Southern States 6.0 5,5 4.2 3.6 o5 Minas/Esp. Santo 6.9 6.3 405 4.3 .68 Northeast 7.5 7.6 6.3 6.1 o80 Frontier States 7.1 701 n.a. 508 o82 Brasil - Total 6.3 5.8 4.4 4,3 o74 Urban 4.7 4.6 3.6 3.6 .77 Rural 7.7 7.7 6.4 6.4 .83 Source: National Academy of Sciences, 19829 from census and POAD da2ao 1980 estimates from Fundacap IBGE, not yet publishedo N\ote that years refer to date of census or PNADo When fertility rates are declining, rates that are estimated by indirect methods reflect the experience of 2-3 years prior to interviews. Neither Colombia nor Mexico is characterIxed by as greei; a degree of regional variability in fertility trends as Brazil, thou.gh bo'nT experienced 11 lower rates in the regions in which their capitals are located and in larger urban centers before the onset of rapid fertility decline at the national level. In Colombia (Table 4), total fertility rates were six to eight births per woman in all four of the country's main regions in 1968-69, compared to 4.5 in Bogota (which accounted for about 13 percent of total population). There was also a differential of four births per woman between total fertility Table 4. Total Fertility Rates, Regional Differentials in Colombia, 1968-80. Region 1968-69 1975-76 1980 80/68 Atlantica 7.8 5.1 4.1 .53 Oriental 7.9 5.1 4.0 .51 Central 6.8 4.1 3.5 .51 Pacifica 5.9 4.6 3.3 .56 Bogota 4.5 2.8 2.8 .62 Colombia-Total 6.7 4.4 3.6 .54 Urban 5.2 3.5 3.0 .58 Rural 9.2 6.3 5.1 .55 Source: Ochoa, Ordonez, and Bayona, 1983. in urban areas (59 percent of the 1973 population) and rural areas. Between 1968 and 1980 total fertility declined by nearly half in all four of Colombia's main regions, and by just under 40 percent in Bogota. Reductions 12 ranged from three to four births per woman, except in Bogota, Urban fertility dropped by 2D2 births, while the rural rate fell by 4.1 births. As in Brazil, the aggregate rate fell by more than e;ther the rural or urban rate, indicating that population redistribution contributed to the eccline. In Colcmbia, both rural and urban total fertil4ty started the decacGe at higher levels and ended it at lower levels than in Brazil, with Colc,,biacs rural fertility decline being more than double that of Brazil in tenms of the decline in number of births. Mexico is also more homogeneous than Brazil in tems of regional differentials in total fertility before the onset of rapid declires. There is little evidence of decline in any region until the midY.970s (Table 5). Comparison of total fertility rates derived from 1960 census data (based on reported children ever born by women in their 40s) with rates from mid°1970 survey data suggest that rates may actually have risen during the 1960s, though caution is required since there was underreporting in the 1.960 census, and probably higher underreporting in regions (Golfo, Sureste, Pacifico Sur) with large Indian populations. The survey data also show a rural-urban differential of 106 births per woman in the early 1.970s, with the rate for the metropolitan area of Mexico City being quite close to that of the urban population in general. During the 1970s, total fertility dec Aned by 30 percent in urban areas and 26 perce,t in rural areas0 In tY'vs, Mexico contrasts with Colombia, where rural declines were greater in zbsolute and relative terins. On the other hand, the onset of fertiHty dec2ire was later in Mexico, and urban areas appear to have led during the earlVnr phases of fertility declines in Colcmbia and Brexil. Also in contrast tc Colombia and Brazil, MexicoIs decline was concentrated in the )ast part o The 2170s, and does not appear to have slowed down recently. 13 Table 5. Total Fertility Rates, Regional Differentials in Mexico, 1960-79. Region 1960 1971-73 1977-79 78/72 Noroeste 6.8 6.3 4.5 .71 Noreste 6.0 5.6 4.1 .73 Norte 6.9 7.3 5.6 .77 Occidental 6.6 7.5 5.5 .73 Centro 6.1 6.0 4.3 .72 Golfo 5.2 6.6 5.7 .86 Sureste 6.2 6.0 4.4 .73 Pacifico Sur 5.8 6.8 5.7 .84 Mexico--Total 6.6 6.6 4.5 .68 Urban n.a. 5.8 3.8 .70 Rural n.a. 7.4 5.5 .74 Source: 1960, Mexico, CONAPO, 1983; 1970s, Mexico, Jefatura, 1981. The Proximate Determinants of Total Fertility. While social and economic changes as well as population programs have played major roles in these fertility declines, in analyzing their effect it is important to recognize that the reproductive behavior of individual couples lies beneath aggregate rates. Biological and demographic factors as well as socio-economic forces influence this behavior, in particular the four proximate determinants identified by Bongaarts (1982): marriage patterns, 14 contraception, lactational infecundability, and abortion0 Few Latin American countries have detailed enough statistical information on reproductive behavior to provide a complete account of the declines that have occurred since 1965. Colombia and Mexico participated in the World Fertility Survey during the mid41970s, and both conducted contraceptive preva-ence surveys later in the decade, but Brazil has yet to conduct a national fertility survey. Despite data limitations particularly information for establishing benchmark estimates to measure the components of declines, it is possible to piece together a fragmentary account of the proximate determinants of recent fertility declines. Assessment of the effects of age at entry into unions and the percent of women in unions on fertility is complicated in Latin America by the influence of consensual unions on reporting practices. All three countries include consensual unions among their marital status categories. BrazilPs 1960 census reported 5 percent of unions as consensual among married women between the ages of 15 and 49; the corresponding figures for Colombia (1964) and Mexico (1960) were 9 and 10 percent0 The problem is not that consensual unions exist but in the likelihood that they are underreportedo3 The data (Table 6) suggest that marriage patterns have changed little since 1960; however, the strength of this conclusion depends cn whether the reporting of marital status improved over the last two decades, particularly with respect to underreporting of consensual unions. If reporting did not change, then one can argue that whle consensual unions were underreported, the bias is the same at both dates and age at marriage and the proportion of women in unions has not changed0 This leads to the important conclusion that fertility declines that have occurred since the mid°1960s have been declines 15 Table 6. Trends in Marital Status and Marital Fertility. MEASURE BRAZIL COLOMBIA MEXICO Percent Single, Ages 20-4: Early 1960s 43.7 45.6 35.8 Late 1970s 51.6 48.3 37.8 Percent Single, Ages 50-4: Early 1960s 9.0 19.8 8.9 Late 1970s 7.8 12.9 5.1 Singulate Mean Age at Marriage: Early 1960s 22.1 21.2 21.3 Late 1970s 23.3 22.1 21.7 Consensual Unions/Percent of Total Unions for Women 15-49: Early 1960s 5 10 9 Late 1970s 5 12 8 Total Marital Fertility Rate: Early 1960s 9.3 11.4 9.7 Late 1970s 7.1 7.4 7.6 Bongaarts' Index (Cm): Early 1960s .67 .58 .67 Late 1970s .63 .57 .64 Sources: 1960s, national population censuses; 1970s, WFS and PNAD survey data. in marital fertility, or more accurately, declines in the fertility of women in unions. This is admittedly a less clear-cut delineation of entry into unions than demographers have in mind when assessing the effect of age at marriage on the time during which women are "at risk," assuming that most births occur in unions. In fact, the majority of births do occur in unions. The three remaining Bongaarts' indices measure the fertility reducing 16 effects of the proximate determinants of fert'lity of women in unions. The first relates to postpartum amenorrhea, whose duration is linked to the prolongation of breast-feeding after childbirth. In most Latin American countries, including Brazil, Colombia, and Mexico, the average duration of breast-feeding is short (less than 10 months), particularly among urban women, and the fertility inhibiting effect of postpartum amenorrhea very limited. Colombia's 1976 tJFS reported that 91.5 percent of urban women and 95 percent of rural women breast-fed their penultimate child. But the median duration of breast-feeding was 7.5 months, with shorter durations being reported for lower order births.4 In Mexico's 1978 WFS, the corresponding figures were 83.3 percent for urban women and 91o9 percent for rural women, with a longer average duration of breast-feeding, 9.8 months. National level data are not available for Brazil, but CPS data for a number of states indicate that the practice is very limited. Median duration of breast-feeding ranged from less than one month in the urban Southeast to around 9 months in the rural Northeast, with low durations (4-5 months) reported in all of the urban areas for which data were available. In the absense of baseline data, most of what can be said about trends is speculative. Differences in the prevalence and duration of breast-feeding by birth order and education suggest that practice is more limited among younger, more educated women, at least until the recent revival of interest stimulated by evidence of its health benefits. Urbanization and increases in female labor force participation may also have had a dampening effect on duration, to the extent that work outside the home requires long periods of physical separation of mothers and infants, and as a consequence of inhibitions imposed by working conditions and employer attitudes0 It is poss,bble, though difficult to document, that declines in breast-feeding contributed to the 17 persistence of high fertility (or even increases in rate) up to the mid- 1960s. Potter and Alba (1983) have suggested that increases in total fertility in Mexico during the 1960s were related to decreased breast-feeding, with rapid rural-urban migration contributing by shifting women from an environment in which it was practiced to one in which it wasn't. The third, and statistically most elusive proximate determinant of fertility is induced abortion. Despite legal strictures, the practice is believed to be widespread in Latin America. A high proportion of abortions are clandestine, known only in terms of the large number of women seeking medical assistance for complications arising from aseptic procedures. Annual hospitalization rates on the order of 40 per 1000 women of reproductive age have been reported for Mexico City (Ordonez 1975, Tietze 1981). In 1974, the International Planned Parenthood Federation estimated that there were about 5 million illegal abortions per year in Latin America (Tietze), which corresponds to an abortion ratio of 300 per 1000 known pregnancies, and a reduction in fertility ranging from 0.8 to 0.9 according to the Bongaarts index (the reason for a range rather than a point estimate is that the index depends on the level of fertility and the proportion of women using contraception). One rationale for government support of family planning programs in the region has been to reduce illegal abortion, and it is possible that recent increases in contraception have partly substituted for abortion as a means of controlling fertility. Most of the statistical evidence about proximate determinants of fertility points to increased and more effective use of contraception as the main reason for the acceleration of fertility declines. In evaluating this evidence, account needs to be taken of the point made in the previous section about abortion. If abortion rates have increased, or even if they have 18 remained constant instead of being reduced by tne subs-itution of contraception, then abortion could be playing a more important rove visa-v¢vis contraception than the statistical evidence suggesCs. Data from WFS and CPS surveys taken in the late 1970s ant' zrly 1980s suggest that very substantial increases in the use of efFpec¢;i ve..ans cf preventing conception have occurred during the last decade. A summary of recent data on contraceptive use is shown in the Table 7. Table 7. Contraceptive Prevalence by Method for Currently Marri,. Howen of Reproductive Age. BRAZIL COLOMBIA MEXICO (A) Sao (b) South (c) Worth- (?1979) (1979) Paulo (1978) (1981) east (1980) ALL METHODS 63.9 65.6 37.3 4601 37.8 IUD - 1.2 04 7.6 8.5 ORALS 27.8 33.5 12.9 1702 1205 SURG-STER* 16.1 14.1 1400 800 8.9 OTHER 20.0 16.8 100 1303 7l8 BONGAARTS' INDEX (Cm) o37 o37 .64 o58 o63 SOURCE: Brazil 9 Thome et al 1982; Colcmbia, Ochoa et al ?.983; Mexico, Ilfatina 1981. *Note: SURG-STER=surgical steriVzation; XUD included in :;o>:,er:1 for Sao Paulo and injections included with IUD in Mexicoo As indicated above, comprehensive national level data cin csontraceptive use are not available for Brazil. However, a sampling of sie ievel daia from CPS surveys gives a sense of the range of regional varizat oi. ,hfe firs§ Brazilian CPS survey conducted in Sao Paulo showed that 5 0ri Marr°ed women of reproductive age in the state were current users of ontraceptMon0 19 Total fertility in Sao Paulo at the time was about 3. While oral contraceptives accounted for the highest percentage of current users, 16 percent of women were surgically sterilized, suggesting that preferences for sterilization in the United States and other industrialized countries has spread to Brazil. Sterilization also plays a major role in other regions of Brazil accounting for 14 percent of women in the South (Parana, Santa Catarina, and Rio Grande do Sul) and Northeast (represented by data from Bahia, Pernambuco, Rio Grande do Norte, Paraiba, and Piaui). In some Northeastern cities, as many as 25 percent of women reported sterilization. Recent CPS surveys for Colombia and Mexico show national averages of 46.1 and 37.8 percent of married women of reproductive age using any method; total fertility rates at the time were 4.2 and 4.7, respectively. Oral contraceptives accounted for the largest share of women using any method, followed by "other." In both, surgical sterilization accounted for a lower share than in Brazil. These data suggest that increased contraception played a significant role in declines in total fertility rates. Assessment of the extent of their contribution requires benchmark estimates of contraceptive prevalence, preferably in the mid- to late 1960s. Unfortunately, evidence for that period is much more fragmentary. Rio de Janeiro, Bogota, and Mexico City were included in a series of surveys conducted in 1964 by CELADE and the University of Chicago, which indicate that contraception was limited both in its extent and its effectiveness (CELADE-CFSC 1972). Current use in the main metropolitan areas was 20-30 percent, and consisted mainly of traditional methods (1972). More recent Colombian surveys show: (1) increases in overall contraceptive use between the late 1960s and middle 1970s, but a slowing of those increases after 1976; (2) narrowing of rural-urban differences, with 20 rural areas apparently accounting for all of the increase in contraceptive use after 1976; and (3) a shift toward more effective methods, particularly in urban areas. That shift helped to offset some of the effect of the slowing of contraceptive use on fertility decline (Colombia, CCRP 1982)o For Mexico, Gallegos et al. (1977) estimated that in 1973 11-12 percent of women of reproductive age and in unions were using modern methods (pills, IUDs, injectables, sterilization, condoms and spermicides). Data from Mexico's 1976 National Fertility Survey showed more than a doubling in the proportion of women using a method between 1973 and 1976. Increased use of methods other than the pill accounted for the change: the share of the pill declined from 64 to 37 percent0 From 1976 to 1979, sterilizations accounted for the largest proportional share of the 11 percentage point r,se in total use and continued to increase in importance between 1979 and 1982. Despite the lack of comprehensive benchmark data on which to base analysis of trends in contraceptive use, a fairly strong case can be made for drawing two conclusions from the data that are available: (1) that contraceptive prevalence rates observed at the end of the 1970s represent substantial increases over rates in the mid1960s, with Mexico laggirg behind Brazil and Colombia in the spread of contraceptive use; and (2) t'hat increased use of more effective methods accounted for most of the gain, ',zading to an increase in the weight of those methods in the overall mix of methods. Thus most of the reduction in fertility that has occurred can be attributed to increased use of effective contraceptive methods rather than other proximate determinants. The importance of this progimate determinant can be illustrated by calculating 'devels of total fertil`ty corresponding to different levels of contraceptive use and other proximate determinants using Bongaarts' framework and values of the four ii.dices (marr-* ge, postpartum 21 amenorrhea, abortion, and contraception) that roughly approximate the experiences of Brazil, Colombia, and Mexico. Table 8. Illustration of Effects of Changes in the Proximate Fertility Determinants on the Total Fertility Rate. Total Marital Fertility 9.18 9.18 9.18 9.18 9.18 9.18 9.18 Rate (TMFR) Index of Amen- .85 .87 .88 .90 .92 .93 .95 orrhea (Ca) TMFR x C a 7.80 7.99 8.08 8.26 8.45 8.54 8.72 Contraceptive Use (percent) 10 20 30 40 50 60 70 Contraceptive Effectiveness 50 57 63 70 77 83 90 (percent) Index of Contra- ception (Cc) .95 .87 .79 .69 .58 .45 .31 TMFR*Ci*Cc 7.38 6.99 6.40 5.72 4.87 3.86 2.67 Total Abortion Rate 1.00 1.00 1.00 1.00 0.83 0.67 0.50 Index of Abortion (Ca) 0.94 0.93 0.92 0.90 0.89 0.89 0.87 Implied TFR TMFR*C *Cc*Ca 6.94 6.50 5.87 5.15 4.33 3.43 2.33 Table 8 summarizes these assumptions and their effect on levels of the total fertility rate. Assuming that the index of marriage remained constant at 0.6, the total marital fertility rate is 9.18 if the other proximate determinants have no fertility reducing effect. Assuming that the index of amenorrhea increases from 0.85 to 0.95 during the first stages of a fertility decline implies a reduction in its impact. Contraceptive use increases from 22 10 percent to 70 percent of women. At the same time the effectiveness of the method mix increases from 50 to 90 percent0 The table also assumes that the total abortion rate remains constant at a level 100 abortions per woman untfl contraceptive prevalence reaches 40 percent, and then declines to 0050 The initial level of abortion is assumed on grounds that less effective contraceptive methods are employed during the initial stages of increased contraceptive use, but are replaced by more effective methods later on, so that the need for abortion declines. With these assumptions, the implied TFR ranges between values of 6.7 when contraceptive prevalence is 10 percent to 2.1 when it is 70 percent0 Intermediate values in the table cover the range of experience observed in data for the three countries. With contraceptive prevalence ,n the 10-20 percent range, TFR varies between 6.3 and 6.7, which is approximtely the level observed in the early 1960so With prevalence at 50 percent, the TFR is 4.1, approximately the levels observed in the three countries in the early 1980s0 Explanations Does anything approaching a common "story" emerge from the ?ertility decline experiences of these three countries? Analysis of the proximate determinants indicates significant change in at least one, and pcssibly two of Coale's three conditions for fertility decline access to the means of fertility control and/or motivation to use those means. On the s&pply side, the data suggest that before 1970 access to contraception .b all 'three countries was limited to the private sector: private fa by planning organizations, private physicians, and commercial sales ( With, or more often without a physician's prescription), and that the marrket" for fertility control (with the possible exception of abortion) was a very restricted one. 23 After 1970, access to modern means of fertility control increased significantly, though the way in which it increased differed in the three instances. In Brazil, the private sector accounted for most of the expansion in access to contraception during the 1970s. Commercial sales of pills played a major role. Sterilizations by private physicians also contributed, particularly in the last half of the decade. State level government intervention, in cooperation with the private family planning agency, promoted the spread of contraception in poorer states in the Northeast. In Colombia, public sector intervention came in the early 1970s, adding to, and eventually taking over the momentum that privately organized family planning programs had created during the late-1960s. Simulation exercises aimed at identifying the role of various sources of fertility control in fertility declines that occurred in Colombia between 1964 and 1975 suggested that public and private organized efforts contributed about equally to the 44 percent of overall decline in fertility that was attributed to these sources (Comite de Trubajo 1976) Commercial sales and private physicians continued to play a major role even while organized programs added to the overall supply. Recent survey data show that the share of organized sources declined during the last half of the 1970s. Public sector intervention appears to have had the greatest impact on increasing access to family planning in Mexico, where access through the private sector (commercial and organized) had been weaker to begin with by comparison with Colombia and Brazil. Most of the increase in contraceptive prevalence tracked by surveys taken during the 1970s can be attributed to public sector programs. The Mexican experience is still too recent to determine whether a gradual shift to commercial sales and other private 24 sources may be occurring. Is supply the key to the acceleration of fertility decline, as suggested by hypothesis A? If there was arn unmet demand for family planning in the late 1960s, increased availability through a variety of sources would appear to have satisfied a substantial part of that demand during the 1970s: providing access to fertility control in instances where it was lacking or substituting cheaper and/or more effective means in instances in which traditional methods were known or available but not widely utilized. Survey data on knowledge and use of contraceptives in the capital cities during the mid-1960s suggest a large gap: in Rio de Janeiro, 93.5 percent of married women in reproductive ages knew of a method, but 32 percent were then current users; in Bogota the corresponding figures were 76 percent and 27 percent, and in Mexico city they were 77 percent and 25 percent (CELADE-CFSC 1972). tJorld Fertility Survey data for Mexico showed a large gap between knowledge (90 percent) and current use (30) percent for the country as a whole as late as 1978. Colombia's 1976 WFS showed 96 percent of women knowing about contraception and 42 percent currently using (Mamlouk 1982). Another measure of unmet demand for family planning in the WFS data is the percent of married women in child-bearing ages who want no more children: 61 percent of Colombian women wanted no more children9 compared to 42 percent who were contraceptive users; n Mexico, 57 percent wanted no more children, compared to 30 percent who were current contraceptors (Kent and Larson 1982). The lower level of unmet demand for family planning in Colombia is a reflection of its earlier steps toward wider access to the means of fertility control. The main problem with hypothesis A is that it does not exclude the possibility of changes on the demand side. In the experiences of these three countries, what suggests that such changes may also have been occurring is the 25 combination of (1) large socioeconomic class differentials in fertility and contraceptive use prior to the recently accelerated declines, and (2) the fact that the acceleration reflected/was made possible by the spread of fertility decline to regions and groups that had not experienced it earlier. In addition to the regional and rural-urban differentials reported earlier, substantial socio-economic class fertility differences also characterized these countries before the onset of rapid aggregate level declines. Baseline data on class differentials are limited, but suggest that they may have increased prior to the onset of aggregage declines and narrowed since. Welti (1982) reports increasing educational class differentials in Mexico City between 1961-63 and 1973-75, which was the decade before the onset of national level decline in Mexico. Survey data for Colombia showed a 3 birth per woman differential in total fertility in 1967-68 between women with incomplete and completed primary education (Birdsall, 1979). By 1976 this differential narrowed to 2 births. More recent (1980) data are not exactly comparable because of differences in their educational classification, but age standardized differences in the number of births per woman appear to have narrowed to 1.2 births in a similar educational breakdown. In Brazil, the differential in total fertility between urban women with no education and those with 1-4 years of schooling narrowed from 2.2 births in 1970 to 1.6 in 1976 (Merrick and Berquo, 1983). Between 1970 and 1976, the greatest percentage decline in total fertility in Brazil occurred among lower income urban women. What accounted for the persistence of high fertility among poorer and less educated classes and regions of these three countries until so recently? A basic point is that the development model which they followed between 1940 and 1970 offered few opportunities to the rural and urban poor 26 for achievement of upward economic mobility through smaller faiiilies. This model emphasized import substitution, with detrimental effects on agricultural development and income distribution because of the policies adopted to promote industrial development in the manufacturing sector. 6 In agriculture, the majority of farm families had plots of land that were too small to support them. With little access to land or credit, their main economic opportunities were in off-farm labor, a risky proposition For the lone individual, but viable for the family unit t^hat could rely on some members (more often women and younger children) to maintain the farm household while others worked off the farm. For the marginalized urban population, the family was also an important economic unit, both in securing jobs through family connections and in broadening the family economic base by having family members work in the informal sector. Neither situation, especiaily when combined with cultural attitudes about married women's economic activity outside the home, provided strong incentives for smaller rather than larger families. Migration provided another outlet for pressures associatecl with rising population growth rates during the 1950s and 1960s. To the extent that migration provided an alternative demographic response to these pressures, it probably delayed the onset of fertility decline in rural areas. But it was also contributing to the eventual downturn in overall fertiPity rates, because it reduced the weight of higher rural fertility in nationEl averages. Migration also fed the growth of the lower income urban populations among whom the spread of fertility control during the 1970s contributed so substantially to recent fertility declines0 All three countries have experienced very substantial rates of rural.-urban population edisriution sir-.z 1950, and cities have absorbed most of their population increase during the last 30 27 years. Between 1950 and 1980, the urban share of the total population increased from about one-third to over two-thirds. Even so, shifts in the geographic distribution of the population account for only a small share of recent declines.8 More important are the changes that have been occurring within the different regions, particularly among the growing lower class urban population. While fertility declines were also occurring in rural areas, it was the urban groups whose weight in the overall total was increasing and without whose participation recent declines would not have been so rapid. Changes in social and economic status of those groups could have contributed to increased demand for fertility control, possibly as the effect of social mobility (changes in the proportion of women in childbearing age belonging to education and/or income classes with higher fertility versus those with lower fertility, as suggested earlier in hypothesis B1) or as an effect of changed socio-economic conditions on these groups, so that in addition to redistribution among groups, there could also have been pressures leading to changes within groups, as suggested in hypothesis B2. Limited available data on fertility declines by socio-economic class suggest that there are elements of between group mobility as well as within group changes in all three cases. While there does not appear to have been a marked improvement in income distribution in any of the three countries between 1960 and 1980, there have been reductions in the proportion of women with no education, and overall increases in average educational attainment, with greater change in Brazil and Mexico than in Colombia. Table 9 is a compilation of data on the educational attainment of women in early childbearing ages during the early 1960s and late 1970s. As far as the data permitted, women in their twenties were grouped into three education 28 categories -- none, 1-5 years or primary level, and 6+ years or secondary level with exceptions noted in the table's footnotes. The most striking shift between 1960 and 1980 occured in Brazil, where the proportion with no education exceeded 40 percent in 1960 ane dropped to 18 percent in 1980, while the secondary (6+) class increased from 9 to 37.5 percent. Mexico also had a comparatively high proportion (36 percent) in the "none" class in 1960, and experienced a decline to 18.7 percent in 1976. At the same time, Mexico shows a slight decline in the secondary (6+) class, probably because its (as well as Ooiombias) 1976 survey relates to married women. They generally have lower average educational attainment than Table 9. Distribution of Young Women by Educational Attainment. Education Brazil Co"o,mbia Mexico 1960 1980 1964 1976 1960 1976 None 40.7 18.0 28.2 16.6 35.7 18.7 Primary (1-5) 50.2 44.5 61.7 6008 41.3 61.1 Secondary (6+) 9o1 37.5 10.1 22.4 23.C 20.2 Brazil: 1960 and 1980 census data on women 20Q29 by years in school. Colombia: 1964 census data on women 2029; 1976 survey data on married women 25-34. Mexico: 1960 census data on women 15Q29; 1976 survey data on married women 24-340 unmarried women, who delay marriage for educational purposes. A slightly older age break (25-34 rather than 20-29) was used to compensate 1cr this, but that also leads to understatement since those women are less like;y to reflect recent increases in educational attainment. Colombia had the hichiest average level of educational attainment in 1960, with only 28 percent of women in the none category, and shows less change between 1960 and i976 than Brazi) or 29 Mexico, though the none group did decline to just below 17 percent, and the secondary category increased from 10 to 22 percent. Comparing the three countries, Brazil appears to have presented the greatest opportunity for distributional changes to have contributed to fertility decline, since it had the highest initial proportion of women with no education. Columbia, with a lower initial level of women with no education, offered the least. Experience confirms that redistribution did, in fact, play a larger role in Brazil. For example, changes in the educational compostion of the population accounted for 55 percent of the decline in children ever born among married urban women in Brazil between 1970 and 1976 (Merrick and Berquo 1983). A similar exercise for Colombia indicated that only 20% of declines between 1968 and 1974 could be attributed to the effects of changes in socio-economic characteristics (Comite de Trubajo 1976). Data were not sufficient to permit similar calculations for Mexico; however, data on fertility declines for roughly similar socio-economic groups in prevalence surveys suggest the Mexican experience has been closer to Colombia's than to '9 Brazil's Mexico is an interesting story in that the educational redistribution that occurred between 1960 and 1976 had little immediate impact on fertility, since the main fertility declines occurred after 1976. In a sense, this is also true of Colombia, since educational attainment was high compared to Brazil before the onset of Colombia's fertility decline, suggesting that in both Colombia and Mexico changes in the educational distribution were not sufficient conditions to trigger rapid declines in fertility. The existence of large difference in total fertility by educational level before the onset of declines suggests that more educated women had comparatively better access to fertility control as well higher motivation to practice it. 30 What is the significance of these changes in femrna educational attainment? Do they add up to a Latin AmericGan variant of the ft=graphic transition, with its own special type of "modernization° cl' reproductive behavior? Evidence that fertility decline during the early 197's was greater within lower income and education groups suggests that '.inks between reproductive patterns and changing socio°economic conditions w@re more complex than linear "progress" as suggested by the modernization model. While increased education changed the balance between the costs and benefits of children at the household or "micro" level, there were also "macro" level social and economic changes affecting the conditions under wh°ch micro leve' choices were being made, either by broadening or limiting those choiceso The generally greater impact of within group changes raises the question of what made smaller families appear more advantageous to lower income and education groups after 1965 than before. Some aspects of the modernizatior argument in demographic transition theory are evident even if they are not labeled as such. For example, increased educational attainmtn49 exposure to the mass media, and availability of new goods (from appliances to bluejeans) raised the consumption aspirations of low income groups, particuTJarly in urban areas 10 Such changes represented increased participation of t:2e population in the modern economy ("modernization"'). These changes were not neutral with respect to fertility and family planning versus other types of behavior, nor did they guarantee equal opportunity to realize new aspirations0 In particular, two sets of societal/institutMonal level cha..geS hat were occurring during the late 1960s and early 1970s warrant atten2t°n. One set consists of the forces, already described, changing access to .nd attitudes toward family planning through expansion of commercial distrib:ion systems and organized family planning efforts0 Promotiona1 activities cf the latter 31 often influenced reproductive attitudes even if they were not the source of supply for a given user. The net effect of these changes was to reduce the monetary, time and psychological costs of controlling fertility. The other set consists of the combined impact of persistent income inequality and inflation, which put low income families in an economic squeeze, making it difficult for them to make ends meet, let alone realize increased consumption aspirations. To put this in perspective, one needs to recall the 1950s, when the momentum of postwar import substitution sustained the growth of the larger cities that were attracting increasing numbers of migrants from smaller towns and rural areas. Real incomes of the urban working classes were growing, in part because price controls on food and other basics (public transport, for example) helped to maintain the buying power of their modest wages, and also because their growing political and economic power helped enabled them to maintain their relative economic position. By the early 1960s, import substitution lost momentum in Brazil; Colombia's and Mexico's problems appeared somewhat later.11 The roots of their problems were political as well as economic. Inflation, fed by the wage demands of urban working classes, played havoc with domestic and international economic relationships. Food production stalled, partly as a result of the disincentive effect of the price controls that had been aimed at keeping prices low for urban workers. There has been considerable discussion and debate about long term trends in real wages in all three countries. Reliable empirical evidence on the purchasing power of low and middle income urban households is very limited. All three have set legal minimum wage levels for several decades. However, wage levels have been changed at irregular intervals and coverage under the legislation has been not consistent with respect to particular population 32 CD~~~~~~~~~~~~~~~~~~-W 0~ 0 EgX Pij LO Ca b M 0 cx ~ T- A~~~~~~~~ D3~~~~~~~- CA CA 2J c X \ a en|f;XkW 2?9/ ffi a f X W 8o~~~~~0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ W iIO a ~~~~~*******t****** IO $ I' I L I I I I ' I I T r I 33 groups over time, making it difficult to draw firm conclusions from trend data. To the extent that the data are representative of any group, they reflect the experience of unskilled blue collar workers in manufacturing and construction. Low skill service jobs are also reflected in more recent data. Much of the debate revolves around whether and how much the purchasing power of a day's work deteriorated, particularly in reference to the prices of such basic necessities as food. A number of authors have argued that real wages deteriorated: in Brazil, from the mid-1960s through the early 1970s (Bacha, 1979); in Colombia during the early 1970s and Mexico later in the 1970s (Berry and Soligo, 1980; Solis, 1983; Gregory, 1983). Figure 1 summarizes real wage trends from the 1950s through the 1970s. Three year averages of available data on mimimum or blue collar wage indices are presented for each of the three countries. The Brazilian data show increases from the mid 1950s through the early 1960s, followed by steady declines until the early 1970s, when a very limited recovery begins. The turnaround in Brazilian wages occurs a few years earlier than the onset of rapid fertility decline that occurred in the mid-1960s. Colombia experienced rising real wages through the early 1970s, followed by declines during the early part of the decade, which is also the period of acceleration in Colombia's fertility decline. Mexico, in contrast, shows steadily increasing real wages until the late 1970s. The onset of fertility decline in Mexico occurred before the turnaround in real wages. Mexico's oil boom (1974-79) and the expansionary economic policies that accompanied it were instrumental in sustaining the rise is wages for a longer period. Since 1980, Mexico has also experienced declines in real wages. Opinions about real wage trends are far from unanimous. Less pessimistic views on Brazil are found in Morley (1982), for Colombia in Urrutia (1981) and 34 Mexico in Gregory (1983)o Some of their skepticism relates to the welfare interpretation of wage data. Household compositions, in parti cular, mediates the impact of individual level earnings experiences on household income and consumption, since one way to compensate for deterioration irn the real wages of the principle earner is to increase the participation of secondary earners when that is possible.12 This point has a direct bearing on the onset of rapid fertility decline. If the income/inflation squeexe is understood in general equilibrium terms, households adjusted to it by attempting to cut costs, increase income, and reexamine the`r priorities with respect to expenditure patterns. For households that were starting Ao shift their consumption patterns in the direction of "modern" goods mentioned earlier, the costs of children were competing with these aspirations. With fertility control becoming less expensive and more acceptable, one adjtstnent was to delay or decide not to have more children. Another was increased labor force participation by married women, again with potential competetion in the use of time for child care. One reason for the high degree of correlation between education and fertility is that increased education raises the value of women's time and creates opportunities for substituting work and earnings cuts-id" the home for such home-based activities as the rearing of children.13 In most of Latin America, female labor force participation rates have been lc- by comparison with other areas, particularly more industrialized countries0 Kr part, this is a reflection of reporting practices that have understated thc economic role of women. This leads to the reporting of very cw participa-ion rates for rural women, even though a very high proportion of them are engaged in homne based artisan and subsistence agricultural activities0 Increased urbanization initially brought even lower rates, since artisan manufacture gave way to 35 factory production, and few women were employed in the latter. Even with the expansion of urban services, female participation rates remained low in comparison to industrialized countries. Only during the last ten to fifteen years have female labor force participation rates in the region registered an appreciable upwary trend. Data for Brazil, shown in Table 10, illustrate the point. In 1960 Brazil's female labor force participation rates exhibited the typical Latin American pattern identified by Durand (1976): rates that ranged between 20-25 percent for women in their late teens and early twenties, with declining rates at older ages. The same basic pattern emerges in 1970, though rates for women in their twenties increased by about 6 percentage points, suggesting the beginning of a new pattern in which peak rates occur at older ages. At the same time, Brazil's rates were far lower than in the United States, where the overall average was more than double that of Brazil, and higher rates continue at older ages, reflecting the return of married women to the labor force that has been occurring since the end of the Second World War. Patterns that are similar to Brazil are observed in Mexico, which had slighly lower participation rates than Brazil, and in Colombia, where rates were slightly higher (census dates were also 3-4 years later). During the 1970s, Brazil experienced a significant rise in female labor force participation. Overall, rates increased by 10-15 percentage points, but more important, the main shifts occurred in peak childbearing ages. The 1980 rate for women aged 30-44 was nearly doubled that of 1960. It is important to note that the "increase" in female labor force participation in Brazil may reflect changes in reporting practices, since questions about economic activity in the 1980 census avoided some of the biases that led to underreporting of women in 1970 and earlier. 36 Table 10. Female Labor Force Particlpation Rates for Brazil, CoWombia, Mexico, and USAO A. All Women, Census Data: Age Brazil Coombia 1960 1970 1980 1964 1973 15-19 23.4 24.2 31.2 21.8 2S58 20-29 20.6 26.i 37.7 23.9 31.1 30=44 17.1 21.2 34.2 19.8 22.8 45-64 14.2 15.8 30.0 17.1 17.A Age Mexico USA 1960 1970 1960 1980 15419 20.7 20.9 27.6 52.9 20-29 19.9 21.3 40O0 68.9* 30=44 17.1 15.9 40o3 64.7* 45-64 17.8 15.6 40.6 50.6 B. Currently Married Women, Survey Data: Age CUo omRi M- exR-- 1976 1976 <25 18o5 13.5 25-34 28.3 19,9 35-44 28.7 25o2 45+ 27o2 25.2 Sources: Mote: Age groups are 20-24 and 25-44. Durand 1976; '.970 census date; Curtin 19820 37 Recent census data are not available to confirm the presence of similar trends in Mexico and Colombia; however, survey data suggest that this is the case. Since the surveys include only currently married women, their female participation rates are not comparable with census data, which relate to all women. Rates for married women are typically lower than for all women. What is striking about the survey results is that rates for 1976 are higher than earlier observations for all women, and that they exhibit the same tendency toward increased participation of older women observed in Brazil. Others (Potter and Alba 1983, Paiva 1983, Carvalho et al. 1982) have also noted that institutional factors that had favored larger over smaller families among lower income groups grew weaker during the 1970s. Particular emphasis is placed on both the way and the extent to which commercialization of agriculture and increased restriction of work opportunities to wage labor have influnced household resource allocation in rural areas. Potter and Alba suggest that the shift to wage labor in Mexico delayed fertility decline by providing an incentive for large families because more work options were available, but that these incentives dissipated when the rapid post-war expansion of Mexican agriculture lost momentum after 1965. In contrast, Parira argued that the expansion of wage labor in Brazil reduced the advantages of large families because there were fewer opportunities for children to work. Underlying these constrasting examples, if indeed they hold up to further empirical testing, are important institutional differences in land tenure patterns that would have to be investigated. Of particular interest is the question of how women's participation in economic activity outside the home was related to these changes. Male proletarianization might have had different fertility implications depending upon whether married women with 38 children are or are not also involved in the shift to wage 9abor. These arguments help to explain declines in fertility in rural areas as well as migration from rural to urban areas. Conclusions The stories are still unfolding, and data are still beina assembled for the task of providing a more complete explanation of the fertIlity declines that have been experienced in Brazil, Colombia, and Mexico since the late 1960s. An attempt to reach closure on the question would take us well beyond the scope of this paper, which has been an attempt to find common threads in their experiences. The data that we have examined support the vi,ew that both supply and demand factors have contributed to the recent acceleration of those declines. Analysis of the proximate determinants of fertility identified increased use of effective contraception as the main proximate determinant of these declines, though increased access to contraception resu&ited from the growth of a combination of channels rather than a single source. What suggests that demand was also important was that these decines were the result of a spreading of lower fertility to groups which had rot previously participated in declines and whose weight in total population was suostantia! enough to influence national trends. While those weights have been increasing in recent decades as a result of geographic and social mobPiLty, important changes were also occurring in behavior patterns within groups. Lower income and education classes in urban areas, in particular, contributed to the acceleration of fertility declines. These groups have, on the one had, been increasing their consumption aspirations, and, on the other, experiencing increasing difficulty in realizing these aspirations on account of such macroeconomic forces as inflation and persistent income inequaiity0 The paper 39 has argued that one of their adjustrnents to these pressures has been to reduce fertility, which has permitted them to allocate both time and monetary resources to other types of consumption, necessities as well as new goods such as appliances. Its account of the speedup of fertility decline thus focuses on the interrelation between micro-level changes in reproductive attitudes and behavior and macro-level social and economic changes, suggesting that the two sets of changes have played complementary roles in the acceleration of fertility decline in these three countries. Brazil's experience offers the most support for the hypothesis that the onset of fertility decline was linked to increased economic pressures on lower income groups. Real wages fell dramatically during the late 1960s, the period when fertility decline accelerated, and without large scale public or private organized efforts to increase access to family planning. The case of Colombia is more difficult to interpret because the economic squeeze of the early 1970s coincided with the expansion of organized efforts to increase access to family planning. The two clearly had reinforcing effects on the accelerated fertility decline that occurred in Colombia after 1970, with a greater overall impact on fertility since total fertility was higher in Colombia than in Brazil to begin with, but fell more rapidly than in Brazil during the 15 years after 1965. It is interesTing to speculate about how much lower Brazil's total fertility rate would have been in 1980 had Brazil followed the Colombian policy of organized efforts to increase access to family planning. Mexico's experience is the most recent, and evidence about the acceleration phase covers a shorter time period. Data from Mexico's 1982 fertility survey suggest that the decline in total fertility in Mexico has caught up with Brazil and Colombia. Mexico's experience offers least support for the economic squeeze hypothesis, since real wages continued to rise in 40 Mexico until the late 1970s because of the oil boom. Fertility decline was underway by the middle of the decade. While the onset of the decline could have been linked to broader structural changes described by Potter and Alba (1983)9 the decision by Mexican authorities to provide public sector support to organized family planning clearly played a key role in the spread of contraceptive use to braoder segments of the Mexican population. It is possible that the deterioration of economic conditions after t`ne collapse of the oil boom has augmented the impact of the program on fertility (which would make it easier to understand the lower than expected level of total fertility reported in 1982), but it is hard to conclude that the onset cf decline was linked to the economic turnaround rather than the program. The fact that the program and its effects on fertility anticipated the economic crisis and were then reinforced by it may have been the result of good foresight or good luck, but it could bring beneficial effects in Mexice's future efforts to deal with its many internal social and economic problems. 41 KEFERENCES Arriaga, Eduardo. 1970. "The Nature and Effects of Latin America's Non- Western Trends in Fertility," Demography., 7:483-501. Bacha, Edmar. 1979. "Crescimento Economico, Salarios Urbanos, e Rurais: 0 Caso do Brasil," Pesquisa e Planejamento Economic, 9:585-628. Beaver, Stephen. 1975. Demographic Transition Theory Reinterpreted, New York: Lexington Books. Berquo, Elza S. et al., 1977. A Fecundidade em Sao Paulo, Sao Paulo, Editora Brazileira de Ciencias. Berry, R. Albert, 1974. "Real Wages in Columbian Manufacturing and Construction during the Twentieth Century," Research Report #7403, Department of Economics, University of Western Ontario, Canada. Birdsall, Nancy. 1979. "Fertility Decline in a Developing Country: the Case of Colombia," Washington, D.C.: The World Bank, Development Economics Department. Bongaarts, John. 1982. "The Fertility-Inhibiting Effects of the Intermediate Fertility Variables," Studies in Family Planning, 13:179-189. Bongaarts, John and Robert G Potter. 1983. Fertility, Biology, and Behavior: An Analysis of the Proximate Determinants, New York, Academic Press. Cabrera, Yolanda P. 1982. "Efectos del Programa Nacional de Planificacion Familiar sobre Natalidad," in CONACYT, Investigacion Demografica en Mexico, 831-43. Carvalho, Jose Alberto M., Paulo T Almeida Paiva and Donald Sawyer. 1982. "The Recent Sharp Decline in Fertility in Brazil: Economic Boom, Social Inequality and Baby Bust," Working Paper #8, Population Council Latin American and Caribbean Regional Office, Mexico City. CELADE, Centro Latinoamericano de Demografia. 1981. Boletim Demografico, 27. CELADE-CFSC. 1972. Fertility and Family Planning in Metropolitan Latin America, Chicago, Community and Family Study Center, CFSC, The University of Chicago. CEPAL, Comision Economica para la America Latina. 1982. Economia Campesina y Agricultura Empresarial, Mexico, Siglo XXI. Chenery, Hollis et.al., 1974. Redistribution With Growth, London, Oxford University Press. Coale, Ansley J. 1975. "The Demographic Transition," in United Nations, The Population Debate, New York, United Nations, Volume 1, 347-55. - 1978. "Population Growth and Economic Development: The Case of Mexico," Foreign Affairs, 56:415-29. 42 Colombia, Corporacion Centro Regional de Poblacion, CCRP, Minls'~erlo de Salud de Colombia, and Wlestinghouse Health Systems. 1982. Second ContraceptIve~ Prevalence Survey, Colombia, 1980, Bogota, CCRP. Comite de Trabajo para el Estudio del Impacto de la Planificacicn Familiar., 1976., Descenso de la Fecundidad y Planificacion Familiar en Colombia, 1964=1975, Bogota. Curtin, Leslie B., 1982. Status of H(omen: A Comp arative Analysi,s of Twen-ty Developing Countries, * Wgton, . V Y~e P_opu -fation _efir_ence Bureau. Dixon, Ruth B. 1971., "Explaining Cross-Cultural Variations in, Age at Marriage," Population Studies, 25:215-33. Durand, John D. 1975., The Labor Force in Economic Development, Princeton, Princeton University7 Press0 Furtado, Celso., 1976., Economic Development of Latin America, Cambridge, England: Cambridge University Press, n dto. Gallegos, Alfredo et al.,, 1977., "Recent Trends -in Contraceptive Use in Mexico," Studies in Family Planning, 8:197-~204., Garcia y Garma, Irma 0. 1979., "Difereniciales de Fecundidad en M~exico,e Demograpfia y Economia, 13:49=-81., Gaslonde, Sainz S. 1976., "Abortion Research in Latin America," Stu,dies in Family Planning, 7:211=17., Gregory, Peter., 1982., "Economic Development and the Labor Market in Mexico,j' World Bank, Development Economics Department. De Janvry, Alain., 1981., The Agrarian Question and Reformism in Latin America, Baltimore: The Johns Hopkins Unilversit-y Pr-ess., Kent, Mary Medeiros and Ann Larson., 1982., Family Size Preferences: Evidence from the World Fertility Sruveys, 9--- in9T n_T.---e-opuF-- - Reference Bureau., Mamlouk, Maria., 1982., Knowledge and Use of Contrace tion n y,enty Developing Countries, Washington, D.,C.,:- Populatin Reference Bureau. Manautou, Jorge Martinez., ed., 1982., The Deimgra hic Revoluticn in Mexico, 1970-80, Mexico, Instituto Mexicanoq! 9ee_quro Soc4Tv Merrick, Thomas w., and Elza Berquo. YS83., The Determrinants of 8Ta%!l's_Rap'!d Fertility Decline, Washington, D.C.,, Nafiona Vcademy P sTo Mexico, Buro de Investigacion de Mercados S., A., 1979., "Fainily Nlanning in Mexico: A Comprehensive Markteting Study of Aw4areness, Attitudes, and Practices among Consumers and RetLailers,'" Mexico, PROFAMO Mexico, Consejo Macional de Cinencia y TecnoloqiE, CONACYTO 198?. Investigacion Demografica en Mexico, 1980, Mexico, CONACVTO Mexico, Consejo Macional de Problacion, COMAP0O 1982. Miexico Uemografico Brevario, 1980-81, Mexico City, COWAPO., 43 Mexico, Consejo Nacional de Problacion, CONAPO. 1982. Mexico Demografico Brevario, 1980-81, Mexico City, CONAPO. Mexico, Jefatura de Servicios de Planificacion Familiar. 1981. Fecundidad y Uso de Metodos Anticonceptivos en Mexico, Mexico City. Mexico, Secretaria de Programacion y Presupuesto, Direccion General de Estatistica. 1979. Encuesta Mexicana de Fecundidad, Mexico City. Morley, Samuel A. 1982. Labor Markets and Inequitable Growth: The Case of Authoritarian Capitalism In Brazil, Cambridge, England: Cambridge University Press. Morris, Leo. et al., 1981. "Contraceptive Prevalence Surveys: A New Source of Family Planning Data," Population Reports, Series M, #5. National Academy of Sciences, Committee on Population and Demography. 1981. Indirect Techniques for Demographic Estimation, Washington, D.C., National Academy Press, Report #10. _ 1982. Levels and Recent Trends in Fertility and Mortality in Colombia, Washington, D.C., National Academy Press, Report #12. _ 1983. Levels and Recent Trends in Fertility and Mortality in Brazil, Washington, D.C., National Academy Press, Report #21. Nortman, Dorothy. 1982. Population and Family Planning Programs: A Compendium of Data through 1981, New York, The Population Council. Ochoa, Luis Hernando, Myriam Ordonez, and Alberto Bayona. 1983. Perfil Socioeconomic y Demografic de la Poblacion Colombiana, Bogota, Corporacion Centro Regional de Poblacion, CCRP. Ochoa, Luis Hernando, and Myriam Ordonez. 1983. Viente Anos de Mortalidad in Colombia, Bogota. Corporacion Centro Regional de Poblacion. Oechsli, Frank, and Dudley Kirk, 1975. "Modernization and the Demographic Transition in Latin America and the Caribbean," Economic Development and Cultural Change, 23:391-419. Ordonez, B. R. 1975. "Induced Abortion in Mexico City," In Epidemiology of Abortion and Practices of Fertility Regulation in Latin America, Washington, Pan America-Health nOrganization. Paiva, Paulo. 1983. "Fecundidade e Padrao de Vida: A Experiencia Brasileira Recente," Paper presented at Congreso Latinoamericano de Poblacion y Desarrollo, Mexico City. Potter, Joseph E. 1982. "El Uso de Variables Intermedias para la Evaluacion de Datos sobre Fecundidad Reciente," in CONACYT, Investigacion Demografica en Mexico, 649-655. 44 Potter, Joseph E. and Francisco Alba. 1983. "Program Development and the Mexican Fertility Transition," Paper presented at 1983 annua6 fieetings of Population Association, Pittsburgh. Potter, Joseph E., Myriam G. Ordonez, and Anthony R. MeashaO. 9376. "The Rapid Decline of Colombia Fertility," Population and Development Review. 2 :509-528. Rodrigues, Walter. et al., 1981. "Contraceptive Practice and Ccmmunity-Based Program Impact in Northeast Brazil,t' Paper presented at mreetings of American Public Health Association, Los Angeles. Sant'Anna, Anna M., Thomas W. Merrick, and Dipak Mazumdar. 1976. "Xncome Distribution and the Economy of the Urban Household: The Case of Belo Horizonte," World Bank Staff Working Paper, Mo. 237. Solis, Leopoldo. 1983. La Realidad Economrica Mexicana: Retrovision y Perspectivas, Mexico,TSTi-gT71tohEdXition. Thome, Marcio. et al., 19820 "Pratica Contraceptiva em Alguns Estados Brasileiros," Paper presented at bi-annual meetings of Associacao Brasileira de Estudos Populacionais, Vitoria. Tietze, Christopher. 1981. Induced Abortion: A World Review, (4th edition), New York, The Population Council. van de Walle, Etienne and John Knodel. 1980 "Europe's Fertility Transition: New Evidence and Lessons for Today's Developing HorThd," Potltion Bulletin, (Population Reference Bureau), 340 Wells, John. 1977. "The Diffusion of Durables in Brazil and Its Implications for Recent Controversies Concerning Braziiian Development," Cambridge Journal of Economics, 1:259-279. Welti Chanes, Carlos. 1982. "Estimacion del Cambio en el Nivel de Fecundidad de la Poblacion del Area Metropolitana de la Ciudad de Mexico entre 1964 y 19769" in CONACYT, Investigagion Demografica en Mexico, 297-311. Urrutia, Miguel. 1981. "Winners and Losers in Colombia's Recent Growth Experience," Bogota: FEDESARROLLO, Processed. Uthoff, B., Andras and Gerardo Gonzalez C. 1984. t3Fertifity anc iomen's Contemporary Labour Force Participation, Education, and Marta` Status: A Comparative Study of Costa Rica and Mexico," Forthcoming in Fertility in Developing Countries: An Eccnoniic Perspective on Re§ear a o Issues. 45 Footnotes 1. This paper is a shorter, revised version of a background paper prepared for the World Bank's 1984 World Development Report, which was prepared with financial support from the Bank. Several Bank staff members, particularly Ricardo Moran, William McGreevey, and Randy Bulatao, provided valuable comments and suggestions. The author alone is responsible for its views and errors it may contain. 2. See the notes to Table 2. 3. Underreporting may well be linked to fertility, since younger women who began a union as a consensual relationship could formalize it after the birth of a child, while women who were in consensual unions in their peak child-bearing years but have not had children may report themselves as "single" (meaning never married) if the union no longer exists by the time they reach age 40. It is also possible that the pregnancy or birth of a first child provides the circumstance that leads couples to enter into civil and religions marriages as well as consensual unions, since data on the mean age of entry into these unions and the mean age of first birth show that the differences between the two is generally less than nine months. 4. Bongaarts' index of postpartum amenorrhea (ci) measures the effect of breastfeeding on fertility in terms of a ratio that expresses the proportional reduction in total fertility attributable to postpartum infecundability. Index values based on Colombia's 1976 and Mexico's 1978 WFS data are identical -- 0.84. Values derived from Brazilian CPS data range between 0.85 and 0.97, and an attempt to construct a weighted average suggested a level of 0.91 for the mid-1970s (Merrick and Berquo 1983). The fertility reducing effect in these three countries is substantially less than in a number of South and Southeast Asian countries, where index values of 0.54 (Bangladesh), 0.58 (Indonesia) and 0.61 (Sri Lanka) have been reported. 5. For accounts of the import substituting industrialization experience in Latin America and its effects on agriculture and income distribution, see Furtado (1976). 6. Potter and Alba (1983) make this point in explaining why fertility decline was so late in coming to Mexico. 7. De Janvry (1981, p. 91) adds that migration adds to the motivation for having a larger number of births because as older children move away, younger ones are needed to maintain the stock of children working in unpaid family labor. 8. For example, recalculation 1980 fertility rates using 1960 rural/urban weights, Brazil's overall total fertility rate in 1980 would have been 5.1 rather than 4.3, Colombia's 4.1 rather than 3.6, and Mexico's would have been 4.6, compared to the observed level of 4.5 births. 46 9. Uthoff and Gonzales, (1984) carried cut a standardization exercise using 1968 and 1970 census data for Mexico and found very limited compositional effect. Garcia y Garna came to similar conclusions about the effects of changing educational composition. 100 See Wells (1977) for a discussion of the diffusion of consumer durables among lower income groups in urban Brazilo Merrick and 3erquo (1983) present data for inerlaced owership of TV sets among low inccme grcups during the 1970so 11. Brazil's industrial sector expanded at a rate of more than 1.0 percent per annum during the 1950s; from 1962 to 1967 the rate fell to less than 4 percent. During the subsequent recovery (1968-74) -he rate of industrial expansion was well above 10 percent; however, the reel income of industrial workers deteriorated under the wage policy that was adopted0 Mexico experienced more rapid industrial growth rates during the 1960s, followed by a slowdown during the 1970s. Colcmbia2s industrial growth rates did not reach levels attained by Brazil and Mexico. lThile the rate of growth slowed during the early 1960s, the recession was not as great as in Brazil (see Furtado, 1976, p. 152)o 12. An illustration of the way that households in Brazil adapted to the earnings squeeze during the early 1970s is found in Sant'Arnna, Merrick, and Mazumdar (1976). That study found considerable var'ation in the capacity to maintain household income levels according to family size, structure, and stages in the family life cycle. For example, households with small children and those headed by women were less likely to have income supplemented by the earnings of a secondary earner. The infonmal sector provided opportunties for secondary earners in cases where a household was able to rely on an additional member!s work outside the home. 13. The logic of the relation between female educational attainment, labor force participation, and fertility has been spelled out in the extension of micro-economic theory that has been labelled as 'e "nedj home economics." Many Latin American social scientists object to that logic on the grounds that its emphasis on the inverse relation bet^seen risinq education and declining fertility as part of a "modernizaE-on" process puts too much emphasis on rationalistic optimising by individual decision makers, with implications that "the market" is doing its work0 The critics feel that the focus on individual behavior neglects overarching "structural" forces which limit the range of individual cVoIce, and the "optimizing" has welfare implications wich suggests that people are becoming better off in the process, when in fact they are iaving a very difficult time maintaining their living standard, or even experiencing a decline in welfare0 In part, the difference in viewpoints is semantic, since "optimizing" is being confused with "optimal"% n.icro-econcmic theory does not preclude the possibility of optimizing in circumstances that require choosing the best possible allocation of resources when resources are shrinking. World Bank value of these findings and provides Demographic Aspects of valuable insights into possibilities for Migration in West Africa- Publications implementing mass programs for K. C. Zachariah and others needy people in villages throughout of Related the world. Volume 1 Interest Staff Working Paper No. 414. September 1980. 369 pages (including statistical an- Volume 1: Integrated Nutrition nexes, bibliography). and Health Care Stock No. WP 0414. $15. Arnfried A. Kielmann and others This volume provides detailed data Volume 2 suggesting that synergism between Staff Working Paper No 415. September malnutrition and infection is probably 1980. 391 pages (including statistical an- the greatest cause of mortality, mor- nexes, bibliography). The African Trypanosomiases: bidity, and retarded growth and devel- Stock No. WP 0415. $15. Methods and Concepts of : opment in children. In an experiment (These Working Papers are background over a period of four years, villagers suisfrMgaini etAna Control and Eradication in received nutrition care, general health studies for Migraton i West Afnca: Relation to Development care to control infections, or both. Dra- Demographic Aspects, described in this C. W. Lee and J. M. Maurice matic improvements, including a 40%- section.) Here is a practical cost-benefit ap- 50% decline in mortality, a 20% reduc- Economic Motivation versus prahto an age-old problem affecting tion in duration of morbidity, and in- Ct ihs etn proach an d prok a frin creases in height and weight. In addi- Cty Lights: Testing humans and livestock alike, the Afri- tion, detailed information on costs is Hypotheses about Inter- can Trypanosomiases. Descnbes new presented that permits the most com- Changwat Migration in techniques that offer tsetse control plete analysis of cost-effectiveness and Thailand without destroying game animals. program relevant costs and benefits Fred Arnold and Susan H. Also summarizes current research in yet available g this kind of field re-r genetic control, the use of traps and search. The study focuses directly on screens, attractants, and pheromones. practical program implications and Staff Working Paper No 416. September Technical Paper No. 4. 1983 107 pages. ways in which such integrated services 1980. 41 pages (including footnotes, refer- ISBN 0-8213-0191-8 Stock No BK 0191. can be applied under field conditions. ences). $5. The Johns Hopkins University Press. 1984. Stock No. WP 0416 $3. Analzingthe Impact of 288 pages. Experiments in Family Hnalyhzering : rjetLC 82-23915. ISBN 0-8018-3064-8. Stock Planning: Lessons from the Health Services: Project No JH 3064. $24.50. Dxevelop ing World Experiences from India, Developing World Ghana, and Thailand Roberto Cuca and Catherine S. Rashid Faruqee Pierce aWorking Paper No. 546. 1982 Volume 11. Integrated Family A comprehensive review of experi- Safes Planning and Health Care mental efforts in the developing world ISBN 0-8213-0117-9. Stock No WP 0546. Carl E. Taylor and others to determine more effectve ways of $3 To village people, pohticians, and in- providing family planning services ternational health planners, health and The Johns Hopkins University Press, 1978. family planning have always seemed 276 pages (including bibliography, index of NEW to fit naturally together. But in the experiments). early 1960s, when international aware- LC 77-16596. ISBN 0-8018-2013-8, Stock Child and Maternal Health queness of the social and economic conse- No. JH 2013, $19 50 hardcover, ISBN 0- Sehices inandiMa:ea THeat quences of surging population growth 8018-2014-6, Stock No. JH 2014, $8.95 Services in India: Tlte moved family planning into a position paperback. Narangwal Experiment of high priority, some intemational What can primary health care and agencies began to advocate separation Family Planning Programs: An family planning do for women and of family planning from health serv- Evaluation of Experience children in the poor and deprived ices. In intemational policy discussions Roberto Cuca areas of the world? Some of the most the question continues to be impor- specific evidence available today to tant. This volume analyzes this ques- Staff Working Paper No. 345 1979 146 support the benefits of these services hon and provides arguments and evi- pages (including 2 annexes, references). is contained in these two studies, dence to support integration of health Stock No. WP 0345. $5 which represent the findings of re- care and family planning; it outlines search carried out during 1967-74 in the purposes underlying the research Fertility and Education: What twenty-six villages in Punjab, India. in this area; and it proposes policy Do We Really Know? questions regarding the effectiveness, Susan H. Cochrane Members of the research staff, which efficiency, and equity of such an inte- A model identifying the many chan- started with 15 people and grew to gration. n oelshog whichieducation m igy htn about 150 by the end of the project, nels through which education might spent many years working with, and The Johns Hopkins University Press. 1984. sharing the lives of, villagers. The 256 pages. depth of understanding that came LC 83-23915. ISBN 0-8018-2830-9.Stock Prices subject to change without notice from this type of sharing enhances the No. JH 2830. $22.50. and may vary by country. act to determine fertility and a review Health fssues zmd Policies in KSenpy: po kdon and of the evidence of the relation between the Developing Ccmntriks DevEopmena education and the intervening varia- Fredrick Golladay (See description under County bles in the model that affect fertility. Work Pap Studescliting.) The Johns Hopkins University Press, 1979 Staff Working Paper No 412 1980 55 Studies listing.) 188 pages (including bibliography, index). pages LC 78-26070 ISBN 0-8018-2140-1, Stock Stock No WP 0412 $3 No. IH 2140, $6 95 paperback Health, NLitriim, and iFaenly Migratioln in West Afkzia: PRanning in Hlndia: A §zrvey o§ DIemoo phic Asse¢ci Expeirmeenas and Special K. C. Zachariah and Julien Conde lFeetilRay and lts lRegulation in Projeccs The first study of the large-scale move- l3angladeslh Rashid Faruqee and Ethna Johnson menit of people in nine West African R. Amin and Rashid Faruqee countnes. Discusses the volume and Staff Working Paper No 507 1982. 108 direction of internal and extemal flows Staff Working Paper No. 383 1980 54 pages (including references) and the economic and social character- pages (including references). Stock No WP 0507 $5 istics of migrants. Stock No. WP 0383. $3. A joint World Bank-OECD study Oxford nftant and c2hfld MoirtaRliy as a University Press, 1981 166 pages (includ- lDeertninan~ off rFetlIidy: Tihe ing 22 maps, bibliography, index). HiemX&lh Policy RMPRicalions LC 80-21352 ISBN 0-19-520186-8, Stock Fredrick Golladay, coordinating Susan Hill Cochrane and K. C. No OX 520186, $19 95 hardcover; ISBN author Zachariah 0-19-520187-6, Stock No OX 520187, Draws on expenence gained from An illustrative analysis that suggests $8.95 paperback. health components of seventy World infant mortality may be an important Bank projects in forty-four countnes component of a fertility reduction pro- between 1975 and 1978 Emphasizes gram in countries where mortality is the disproportionately high expendi- high and few couples are able to have IapŽln and Famnilly tures incurred on curative medicine, the number of surviving children they P:lanm7n_n, in 3anglzadeshI: A maintenance of expensive hospitals, desire. Sftudy o$ the leseazch and sophisticated training of medical World Bank Staff Working Paper No. 556 Mohammad Alauddin and Rashid personnel at the cost of preventive19344pg.Faue care for the majonty of the people. 1983. 44 pages Faruqee Points out that low-cost health care ISBN 0-8213-0147-0 Stock No. WP 0556. Reviews major studies on family plan- systems are feasible and recommends $3. ning and on fertility trends, profiles, that the Bank begin regular and direct . .and determinants. Evaluates results of lending for health, in addition to hav- En&ega&ng IamiRy Manni such studies and cnhques their meth- ing health components as part of pro- wnth HeaIHL Gezvicn : IDos Ht odology and application. Underscores )ects in other sectors Hellp? need for continued study and suggests jector Poin y other. 1980 90 pages in- Rashid secor . directions for future research to im- Sector Policy Paper. 1980 90 pages (in- Rashdt Farsuqee prove the Bangladesh population cluding 8 annexes, 4 figures, map) Staff Working Paper No 515. 1982. 47 problem. Stock Nos BK 9066 (Arabic), BK 9067 pages. World Bank Staff Working Paper No. 557. (English), BK 9068 (French), BK 9069 ISBN 0-8213-0003-2 Stock No. WP 0515. 1983 176 pages. (Spanish). $5 - $3 ISBN 0-8213-0150-0. Stock No WP 0557 $5 I ! L I L \ \I rak$22 pkOon and Pove:t r.n the K stv= rih s1n vg"u E'alIoplIngWo2ld I i m ji~'~Q©~ f 2 1t$h0o Nancy Bi`dsall Staff Worlking Paper No. 404. 1980. 96 1 d I t g pages (including 2 appendixes, bibliog- raphy) Stock No. WP 0404. $3 Wi g m \n D 6I I P is | ulpk'f.Ln Policy aind 7InmiRy rI @ L$ L_________ PlIanzlIig PIogza s: leEzns in __________ [PM 2Eolilcy 2ndl AdrninistbrafEon .rnl t . > -=Kandiah Xanagaratnam and F 7 ~Tsh xm, Catherine S. Pierce ' ~ ~ L.............i ~~~~~~~ Staff Wor king Paper No. 447 1981. 80 [ J __________________ pages (including bibliography, appendixes). VVICS53col nmS~Z ftoC s M$bMvoStock No. WP 0447. $3. NEW published as World Population Projec- falling mortality, and urbanization- tions: Short- and Long-term Projections by factors that tend to lower fertility else- Short-term Population Age and Sex for All Countries, with Re- where. Calls for a viable populaton Projection, 1980-2020 and Long- lated Demographic Statistics policy and programs appropnate to term Projection, 2000 to 1983 391 pages. the culture. Stationary Stage by Age and ISBN 0-8213-0355-4. $30 paperback. Staff Working Paper No. 559. 1983. 116 Sex for All Countries of the pages. NEW ISBN 0-8213-0152-7. Stock Nos WP World 0559i $5 My T. Vu, under the supervision Rapid Population Growth in of K. C. Zachariah Sub-Saharan Africa: Issues and Regional Aspects of Family This report gives detailed population Policies Rgoa set fFml projections by age and sex for each Rashid Planning and Fertility Behavior country at five-year intervals from 1980 Rashid Faruqee and Ravi Gulhat in Indonesia to 2020 and at twenty-five year inter- No other country has higher fertility Dov Chernichovsky and Oey Astra vals from 2000 to the year in which than Kenya and its neighbonng coun- Meesook population becomes stationary in each tnes in Sub-Saharan Afnca. This country. The implied fertility and mor- Working Paper examines the reasons Staff Working Paper No. 462 1981 62 tality measures are also given. Revised for fertility rates staying high, even pages (including appendix, references). annually The 1984 edition will be rising in the face of greater education, Stock No. WP 0462 $3 "6~ The World Bank L- Publications Order Form SEND TO: YOUR LOCAL DISTRIBUTOR OR TO WORLD BANK PUBLICATIONS (See the other ttde I/ thoA lrtmt ) P.O. BOX 37525 WASHINGTON, D C 20013 U S.A. Date Name Ship to (Enter if different from purchaser) Title Name Firm _ Title Address Firm City State_ Postal Code Address Country Telephone j City _ _ State_ Postal Code Purchaser Reference No Country Telephone Check your method of payment Enclosed is my D Check 1 International Money Order O Unesco Coupons l International Postal Coupon Make payable to World Bank Publications for U S dollars unless you are ordering from your local distributor Charge my O VISA 1l MasterCard El Amencan Express E Choice (Credit cards accepted only for orders addressed to World Bank Publications) Credit Card Account Number Expiration Date Signature Ol Invoice me and please reference my Purchase Order No Please ship me the items listed below Customer Internal Stock Number Author/ Title Routing Code Quantity Lnia Priir fiital Amount 1 _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ I _ All prices subject to change Prices may vary by country Allow 6-8 weeks for delivery Subtotal Cost S - Total copies Air mail surcharge if desired ($2.00 eachi s Postage and handling for more than two complimentary items ($2.00 eachl S Total S IERo-0053 Thank you for your order. Attn Mr Giancarlo Bigazzi Attn: Mr J MV Hernandez Via Larnarmora 45 Castello 37 Din pmlblicaiCon 50121 Madrid ARGENTINA Florence SRI LANMCA AND T1-1E MALDIVES Carlos Hirsch, SRL, JAPAN Lake House 3ookshop Attn Ms Monica Bustos Eastern Book Service Attn Mr 'victor Walatara Flonda 165 4° piso Attn Mr Terumasa Hirano 41 Wad Ramanavake Mawatha Galeria Guemes 37-3, Hongo 3-Chome, Bunkvo-ku 113 Colombo 2 Buenos Aires 1307 Tokyo SWEDEN AUSTRALIA, PAPUA NEW GUINEA, KENYA ABCE Fritzes Kungl. Hovbokhandel FIJI, SOLOMON ISLANDS, Afnca Book Services (E A ) Ltd Attn Mr Eide Segerback FIJI, SOLOM N ISLANDS,Attn Mir M B Dar Regienngsgatan 1L, Box 16356 WESTERN SAMOA, AND P O Box 45245 S-103 27 Stockholm VANUATUNarbSWTE AD The Australian Financial Review Nairobi §W5TZERtAND Information Service (AFRIS) KOREA, REPUBLIC OF Libraine Pavot Attn Mr David Jamieson Pan Korea Book Corporation Attn Mr Flenri de Perrot 235-243 Jones Street Attn Mr Yoon-Sun Kim 6, rue Grenus Broadwav P 0 Box 101, Kwanghwamun 1211 Geneva Sydney, hSW 20001 Seoul TANZANtA BELGIUM MALAYSIA Oxford Universitv Press Publications des Nations Unies University of Malaya Cooperative Attn Mr Anthonv Theobold Attn Mr jean de Lannoy Bookshop Ltd Maktaba Road, P 0 Box 5299 av du Roi 202 Attn Mr Mohammed Fahim Htj Dar es Salaam 1060 Brussels Yacob THAILANiD CANADA P 0 Box 1127, Jalan Pantai Baru Central Department Store, ktead Office Le Diffuseur Kuala Lumpur Attn Mrs Ratana Attn Mrs Suzanne Vermette MEXICO 306 Silom Road C P 85, Boucherville J4B 5E6 INFOTEC Bangkok Quebec Attn Mr Jorge Cepeda Thailand Management Association San Lorenzo 53-11, Col del Valle, Attn Mrs Sunan COSTA RICA Deleg Benito Juarez 308 Silom Road Librena Trejos030MeioDFBagk Attn Mr Hugo Chamberlain 0310 Mexico, D.F Bangkok Calle 11-13. Av Fernandez Guell MIDDLE EAST TUNISIA San Jose Mliddle East Marketing Research Societi Tunisienne de Ditfusion DENMARK Attn MBureau Attn Mr Slaheddine Ben Hamida Sanfudsltteratur Attn Mr George Vassilou 5 Avenue de Carthage Attn Mr Wilfried Rolotf M.itsis Bl ave Tuni Rosenderns Alle 11 Nicosia TURKEY DK-1970 Copenhagen V C% prus Haset Kitapevi A S EGYPT, Arab Republic of (Branch offices in Bahrain, Greece, Attn Mr Fzzet izerel Al Ahram Morocco, Kuwait, United Arab 469, Istiklal Caddesi Al Galaa Street Emirates, Jordan) Bevoglu-lstanbul Cairo NETHERLANDS UNITED £NiiGDOM AND MBE BV NORTHERN IRELAND FINLAND Attn Mr Gerhard van Bussell Microinfo Ltd. Akateeminen KrlNakauppa Noorderwal 38, Attn Mr Rov Selwvn Attn Mr Kan Litmanen 7241 BL Lochem Newman Lane, P 0 Box 3 Keskuskatu 1, SF-00100 NORWAY Alton, Hampshire GU34 2PG Helsinki 10 Johan Grundt Tanum A S England FRANCE Attn Ms Randi Mikkelborg UNITED STA-E World Bank Publications p 0 Box 1177 Sentrum The World Benk Book Store 66, avenue d'lena Oslo 1600 19th Street, N W 75116 Paris PANAMA Washington, D C 20433 GERMANY, Federal Republic of Ediciones Libreria Cultural P'anamena (Postal address P 0 Box 37525 UNO-Verlag Attn Mr Luis Fernandez Fraguela R Washington. D C 20013, U 5 A Attn Mr Joachim Krause Av 7. Espana 16 Baker and Tavlor Companv Simrockstrasse 23 Panama one I 501 South Gladiola Avenue D-5300 Bonn I PHILIPPINES Momence, Illinois, 60954 HONG KONG, MACAU \ational Book Store 380 Edison )A.av AsHa 2000 Ltd Attn Mrs Socorro C Ramos Reno., Nevada, 89564 Attn Ms Gretchen Wearing Smith 701 Rizal Avenue 50 Kirbv Averue 6 Fl . 146 Prince Edward Road . Manila Somerville, New Jersey, 08876 Kowloon PORTUGAL Commerce, Georgia 30599 INDIA Ili\ raria Portugal ComernnAsc,aGesri 09 UBS Pblishrs' istriutorsLtd\ttrn Mr Antunio Alves MartinsBennAo'ae Rti,i Do Carmo 70-74 9730-E Georg- Palmer Highway UBS Publishers' Distributors Ltd Ru. Do Carmo ,'(-74 9Lan0hamGe,r M arlad, 207rO61h Attn Mr D P Veer 1 ~t ahm Mrln,27 3 Ansarn Road, Post Box 7015 I Blackwell Nodh America, Inc New Delhi 110002 l Fries Mll RoAd (Branch oftices in Bombay, Bangalore, SALDI ARABIA 1001 Fries MNil Road Kanpur, Calcutta. and Madras) lanr Book Store Blackwood, New Jersey 08012 INDC)NESIA Attn Mr Akram Al-Agil Sidnev Kramer Books rt Indira Limited r 0 Box 3196 1722 H Street, N W Attn Mr Bambang Wahvudi Ri%adh Washington, D C. 20006 JI, Dr Sam Ratulangi No' 37 SINGAPORE, TAIWAN, BURMA United Nations Bookshop Jakarta Pusat Intormation Publications Private, Ltd United Nations Plaza IRELAND Attn Ms Janet David New York, N Y. 10017 TDC Publishers 02-06 1st Floor, Pei-Fu Industrial VENEZUELA Attn Mr James Booth Building Libreria del Est7 12 North Frederick Street 24 Ne%% Industrial Road Attn Mr Juen Pencas Dublin I Singapore Avda Francisco de Miranda, no 52 ITALY SPAIN Edlficio Galipan, Aptdo 60 337 Licosa Commissionaria Sansoni SPA Mlundi-Prensa Libros, S A Caracas 1060-A I HG 3881.5 .W57 W67 No.692 1- 1ERRICK, THOMAS WILLIAM. I RECENT FERTILITY DFl_ The World Bank Headquarters European Office Tokyo Office 1818 H Street, N.W. 66, avenue d'lena Kokusai Building Washington, D.C. 20433, U.S.A. 75116 Paris, France 1-1 Marunouchi 3-chome Telephone: (202) 477-1234 Telephone: (1) 723-54.21 Chiyoda-ku, Tokyo 100, JapaX Telex: WUI 64145 WORLDBANK Telex 842-620628 Telephone: (03) 214-5001 0 RCA 248423 WORLDBK Telex: 781-26838 Cable Address: INTBAFRAD WASHINGTONDC E. 0 ISSN02532115ISB 0-813-090n