Childbearing Within Marriage and [629942]

1
Childbearing Within Marriage and
Consensual Union in Latin America, 1980 –20101
Benoît Laplante
Professor, Institut national de la recherche scientifique, Université du Québec, Canada.
Teresa Castro -Martín
Research Professor, Spanish National Research Counci l.
Clara Cortina
Assistant Professor, Universitat Pompeu Fabra, Spain.
Teresa Martín -García
Researcher, Spanish National Research Council.

1 This is an early draft of an article that has subsequently been published in Population and Development
Review. Complete citation informatio n for the final version of the paper, as published in the print edition of
Population and Development Review, is available on Wiley Interscience’ s online journal service, accessible
via the journal ’s website at http://www.blackwellpublishing.com/pdr.

2
Abstract
This article compares the fertility patterns of women in consensual union and marriage in 13 Latin
American countries, using census microdata from the four most recent census rounds and a
methodological approach that combines the own- children method and Poisson regression. Results
show that in all these countries, fertility is slightly higher within consensual un ion than marriage and
that the age pattern of fertility is very similar in marital and non-marital unions. Further analyses
show that over the period considered, childbearing within a consensual union has passed from rare to increasingly common, although not yet mainstream, for highly educated women in most countries
examined. Results show that in Latin America, at least since the 1980s, women’s childbearing
patterns depend on their age and on their being in a conjugal relationship, but not on the legal nat ure
of this relationship. The similarities in reproductive behavior between marital and non-marital unions
are not confined to the socially disadvantaged groups, but apply as well to the better off.
Keywords
Nuptiality, family formation, marriage, cohabitation, consensual union, non- marital fertility, census
microdata, IPUMS, Own Children Method, Poisson regression, Latin America

3
Childbearing Within Marriage and
Consensual Union in Latin America, 1980 -2010
One of the most salient demographic features of L atin America is its dual nuptiality regime (Castro –
Martín 2002). Marriages and consensual unions coexist side by side in all countries of the region,
although the prevalence of consensual unions varies significantly from country to country —from
about 20% of all conjugal unions among women aged 15 to 49 in Chile up to 74% in the Dominican Republic (Castro -Martín et al. 2011) —as well as across regions within the same country (Esteve ,

Lesthaeghe and López -Gay 2012) and across social classes (Arriagada 2002).
Both forms of conjugal union have achieved similar levels of social acceptance, but they often
differ in terms of stability, legal obligations, and economic rights after breakdown (Quilodrán 1999; De V os 2000; Castro- Martín 2002; Rodríguez Vignoli 2004). Un like what occurred in the developed
world, where cohabitation did not achieve social —and statistical—visibility until the 1980s,
consensual unions have been an integral part of the family system in Latin America for centuries
(Socolow 2000). Furthermore, w hereas in North America and many European countries cohabitation
is often a transitional stage leading to marriage, in Latin America the prevalence of consensual unions remains high in later stages of the life course, suggesting that they are often regarde d as a functional
alternative to marriage. Nevertheless, the most notable difference is that, whereas in North America and Europe —with several exceptions such as the Nordic countries, France or Quebec —cohabitation
tends to be a childless stage, in Latin America it is a common family arrangement for bearing and raising children. This feature blurs the differences between de jure unions and de facto unions.
According to a recent study, in the Latin American region, the proportion of births from lone mothers rose from 7% to 15% from 1970 to 2000, and the proportion of births that took place within a
consensual union rose from 17% to 39% ( Castro -Martín et al. 2011). In the 21
st century Latin
America, hence, more children are born outside than within marriage.
This new setting is what has motivated this study. We wish to explore further the similarities and
differences in the reproductive behavior of married women and women living in a consensual union.
We know that for many Latin American women, marriage is not a prerequisite for having children,
but we aim to measure more precisely the differences in fertility patterns of formal and informal
unions over the reproductive age range. We also want to explore whether these fertility patterns by

4
union type are equival ent across women with different levels of education, which we use as a proxy
of social and economic resources and opportunities as well as relative position in the social hierarchy.
In the European and North- American literature, cohabitation and childbeari ng within
cohabitation have been usually discussed from the perspective of the Second Demographic Transition
(Seltzer 2000; Kiernan 2001; Sobotka and Toulemon 2008). The spread of cohabitation is regarded as
a consequence of secularization trends, rising e xpectations of personal autonomy, rejection of Church
and State intervention in the regulation of private life, and growing importance of personal
satisfaction within the couple relationship (Lesthaeghe 2010). Recent studies have also shown that
cohabitati on in some countries can be also related to socioeconomic disadvantage, and used as an
alternative to marriage by people with few economic resources or poor economic expectations (Goldstein and Kenney 2012; Kiernan et al. 2011)
In Latin America, consensua l unions have traditionally been common among underprivileged
social sectors and in rural areas —leading to it being dubbed the “poor people marriage”—, while
they were very rare among the upper class. This pattern suggests that initiating the family formation process within a consensual union rather than a marriage might not always be the result of personal
choice but, at least in part, the consequence of limited economic and social opportunities (Greene
1991; Castro- Martín 2001; García and Rojas 2004). How ever, an important change in nuptiality
patterns has occurred over the last two decades. The presence of consensual unions had started to become noticeable among the well-educated and in urban areas by the end of the 1990s (Parrado and
Tienda 1997). This presence is more manifest and better documented in the Southern Cone ( Cabella,
Peri, and Street 2005; Cabella 2009; Laplante and Street 2009; Binstock 2010). Other studies have
shown that the recent rise in unmarried cohabitation has encompassed all educati onal groups, not just
the underprivileged ( Castro -Martín , Martín -García, and Puga González 2008; Quilodrán 2011).
According to data from the 2000 census round, a significant proportion of Latin American university –
educated women aged 25 to 29 and living in a conjugal union are cohabiting rather than being
married, the proportion exceeding 30 percent in Argentina, Colombia, Cuba, and Peru (Esteve,
Lesthaeghe, and López -Gay 2012).
The fact that cohabitation has recently disseminated among the middle and upper class has led
researchers to make a distinction between “traditional” and “modern” consensual unions (Quilodrán
2011). “Traditional” consensual unions, still the most common in all Latin- American countries, are
generally associated to cultural heritage, l imited economic opportunities, and asymmetrical gender

5
relations. By contrast, “modern” consensual unions, still in an emerging stage, are considered as the
result of a conscious choice in the pursuit of individual autonomy, freedom from institutional cont rol,
and less asymmetry between the genders, pretty much along the lines of the Second Demographic Transition (Lesthaeghe 1995, 2010; Billari and Liefbroer 2004).
Yet there is no complete consensus on the underlying causes and proper interpretation of the
more recent expansion of cohabitation in Latin America. Some researchers consider that it should be
attributed to modernity and the advance of the Second Demographic Transition in the region (Esteve ,
Lesthaeghe, and López -Gay 2012). Other researchers point at additional catalysts, such as the
increasing uncertainty that the middle class is now confronted with in their working, social, and
family life (García and Rojas 2004; Arriagada 2007; Quilodrán and Castro -Martín 2009). Moreover,
there are large differ ences not only in the prevalence of cohabitation, but also in its social meaning,
symbolic value, motivation, and consequences, across countries, social classes, and ethnic groups
(Covre -Sussai et al. 2014). In many indigenous communities, lack of a marria ge certificate is deemed
irrelevant and does not lead to family instability, whereas in the poorest social segments, it is commonly a sign of precariousness, exclusion, and vulnerability. In the middle and upper classes, it
is not clear whether the emergin g form of cohabitation is a step in the union formation process that
precedes formalization, or it is an alternative to marriage and hence a family arrangement in which
children will be born and raised (CEPAL 2002).
This article is an attempt to further ou r understanding of the process of family formation outside
of marriage in Latin America. Comparing the reproductive behavior of cohabiting and married
women constitutes a promising avenue for understanding how consensual unions fit into the family
system (Raley 2001). Recent research has shown an important change in the socioeconomic profile
of consensual couples in the region, due to the recent uptake of cohabitation by highe r educated
young adults ( Esteve,
Lesthaeghe, and López -Gay 2012). Our main objecti ve is to find whether there
is a similar change in the socioeconomic profile of childbearing within cohabitation. By comparing
the reproductive patterns of married women and consensually partnered women, we should be in a
better position to ascertain whether consensual unions are short -lived couple relationships, trial
marriages, “poor people marriages ,” or long- lasting alternatives to marriage. We should also be able
to explore the diversity in meanings of cohabitation across Latin- American societies and across
social classes. We are particularly interested in ascertaining whether the emergent pattern of cohabitation among the middle and upper class corresponds to a trial marriage, where children are

6
postponed until the relationship is formalized, or to an alternative to marriage, where children are
typically born and raised, as it has traditionally happened in the disadvantaged groups.
The prevalence of cohabitation and births within cohabitation in Latin America
In Latin America, consensual unions have be en part of the family system for centuries and nowadays
they coexist side by side with marriage configuring the distinct nuptiality system of the region (Castro -Martín 2002). Notwithstanding their key role in the family formation process, the prevalence
of consensual unions varies considerably across Latin American societies. In countries such as the
Dominican Republic, Panama, Honduras, Nicaragua, Peru, Colombia , and Uruguay —listed in
descending order of prevalence—, the proportion of consensual unions eve n surpasses that of
marriages among women in union of reproductive age (see Table 1). The Dominican Republic is the country with the highest prevalence of consensual unions: three out of four women aged 15 to 49
currently in union live in an informal union. Cuba, El Salvador, Venezuela, Ecuador, Brazil , and
Paraguay also have a relatively high prevalence of consensual unions among partnered women of reproductive age, ranging from 49% to 40%. The lowest prevalence in the region is currently
observed in Chile , where only one -fifth of the unions are consensual, but this level is based on the
2002 census and the still unreleased data from the 2012 census might show an increase.
[Table 1 about here]
According to vital statistics, the number and proportion of non- marital births are remarkably
high in most countries of the region. In the 2000s, the proportion of births from unmarried women
was higher than that from married women in all Latin American countries for which data are
available (see Table 2). In some coun tries, such as the Dominican Republic, Venezuela, Cuba, and
Panama, the proportion of non- marital births reaches four -fifths of all births. In some countries for
which trend data are available, such as Panama or El Salvador, the proportion of births to unm arried
mothers was already very high in the 1970s. In the rest of the countries, there has been a remarkable increase.
[Table 2 about here]
However, vital statistics in many Latin American countries suffer from under -registration
(Harbitz, Benítez Molina, and Arcos Axt 2010) and, with few except ions such as Costa Rica, do not

7
provide information on whether or not the unmarried parents are living together, so children born
from a mother living in a consensual union are not reported separately from those born from a mother
who does not have a co- residential partner.
In order to distinguish births from cohabiting mothers and lone mothers, one need s to resort to
census data or surveys with retrospective birth histories. However, the latter are not widely availab le
for Latin American countries . A recent analysis based on census microdata for 13 Latin American
countries documents that there has been a considerable increase in the proportion of births outside
marriage over the last decades, and that most of this ris e is concentrated in cohabiting unions ( Castro –
Martín et al. 2011). Table 3 illustrates the changing distribution of births in the region according to
the conjugal status of the mother from 1970 to 2000. The dramatic decline in the percentage of births
within marriage over the whole period (from three -quarters to just nearly one -half) goes in parallel
with the significant rise in births from women in consensual unions (from 16.8% to 38.9%). Data show that the percentage of births from lone mothers has also increased, but more modestly (from
7.3% to 15%). In other words, the fact that non- marital births have surpassed marital births since the
beginning of this century in Latin America is mainly due to the increase in births from parents in consensual union.
[Table 3 about here]

Data
We use data from the IPUMS collection of harmonized census microdata files from the four most
recent census rounds available ( Minnesota Population Center 2013) . Census data contain reliable
information on the current conjugal situa tion of all individuals (Rodríguez Vignoli 2011) and provide
a workable alternative to vital statistics or biographical surveys when used with the own -children
method of fertility estimation. We focus on 13 Latin American countries for which such data are
available: Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Cuba, Ecuador, Mexico, Panama,
Peru, Uruguay, and Venezuela. Depending on the country, between one and four censuses were
available and included the information we needed.
Although access to comparable census data for a larg e number of countries covering four
decades is a major advantage, census data also have important drawbacks. First, data pertain only to
women’s current union status and not to union status when giving birth; second, da tes of entry into

8
union are not reported, and hence no union durations can be computed. To minimize possible bias we
focus our analysis on births born in the year prior to the census.
Methods
We compare the fertility of women in consensual union and marria ge estimating age -specific fertility
rates (ASFR s) and the total fertility rate (TFR) by union type. We do not interpret this TFR as an
approximation of completed fertility, which it is not when computed for a time -varying characteristic
such as conjugal s ituation, but as a measure of the overall intensity of fertility within each union type.
We estimate ASFRs and TFR using an approach that combines the own -children method and Poisson
regression, which allows us to compute standard errors and thus test whet her ob served differences
between marriage and consensual union are statistically significant.
Castro -Martín et al. (2011) report that apart from the level of education, in which we are mainly
interested, other markers of social position—labor force status , rural/urban residence, and home
ownership —have a net effect on childbearing outside of marriage. Labor force status could be of
special importance given the recent increase in the labor participation rate of women throughout
Latin America (Abramo and Valenzuela 2005). Given our interest for the recent uptake of
cohabitation among the middle and upper class , we estimate total fertility rates by union type
according to educational level, but also according to these three variables .
Measuring fertility with in marriage and within consensual union
According to Heuveline and Timberlake’s (2004) typology, cohabitation is an alternative to marriage
when individuals choose to cohabit instead of marrying with the intention to form a family as a married couple would. Bearing and rearing children within cohabitation signals that this union type is
regarded as an appropriate family setting. Comparing the fertility patterns of women within marriage
and consensual union, hence, is relevant to ascertain whether cohabitati on has become or is on the
way of becoming an accepted alternative to marriage in a given society. However, comparing the
fertility within marriage and consensual union involves some technical challenges. Fertility is
commonly estimated using vital statistics, and vital statistics generally report whether children are
born to married parents or an unmarried mother, but do not typically report whether the unmarried
mother is cohabiting with the child’s father . Vital statistics are still largely computed following the
traditional distinction between marital and non-marital fertility, and thus have not usually

9
incorporated the social phenomenon of cohabitation into the birth statistics. For this reason, it is hard
to find fertility estimates by union type based on vital statistics (Klüsener, Perelli -Harris, and
Sánchez Gassen 2013). Attempts at estimating fertility within marriage and within consensual union
performed since the 1990s show a variety of solutions, but most of them rely on the use of survey
data wi th retrospective histories (e.g. Do Valle Silva , Henriques, and De Souza 1990; Verdugo Lazo
1994; Brown and Dittgen 2000; Raley 2001; Hoem and Mureșan 2011; Van Hook and Altman. 2013) .
Unfortunately, as we pointed already, the survey data required to make use of these solutions are not
available for all Latin American countries.
Dumas and Bélanger (1998) compared levels of fertility within marriage and cohabitation in
Canada using data from a retrospective biographical survey, but with an approach that can be used with census data. They estimated five -year age group birth rates for the two union forms, for two ten-
year periods, 1975–1984 and 1985–1994, and for two regions, Quebec and the rest of Canada. They computed TFRs for each region and period, based on women’s conjugal status at the time of birth.
They concluded that the fertility of cohabiting women was lower than that of married women in bot h
regions and in both periods, but that those differences were smaller in Quebec than in the rest of Canada. In this analysis, we adapt their approach to census data. Furthermore, combining the own-
children method and Poisson regression enables us to estim ate standard errors and confidence
intervals. See the Technical Appendix for an overview of the own -children method; we explain our
use of Poisson regression below.
Conditional age -specific fertility rates and TFR
Age-specific fertility rates and the total fertility rate are well -established measures of fertility whose
meaning and properties are also well known. They are usually defined and computed for all women
in reproductive age, commonly women aged between 15 and 49. They are sometimes used in the
study of differential fertility and computed for subgroups of women defined by some relevant
characteristic such as ethnic group or place of residence. Although we should be cautious with the
interpretation, t echnically , nothing prevents computing age -specific rates within groups defined by a
time-varying characteristic such as conjugal situa tion: this is exactly what the traditional marital
fertility rate is . Because these rates are computed within categories of a given variable, they are
conditional in the st atistical sense and could be referred to as “conditional ASFR s” and the
“conditional TFR” respectively . Therefore, in our case, the se conditional age -specific fertility rates

10
are the rates of giving birth at a given age while being married, living in a con sensual union, or not
living in a union. The conditional TFR is the sum of such rates and provides an estimate of the
number of children born to a woman continuously married, continuously living in a consensual union
or continuously not living in a union between the ages 15 and 49 in the year for which the rates are
computed. The usual fictitious cohort is broken down into three components that are also fictitious
cohorts. The operation can also be interpreted as a decomposition exercise: the usual age -specific
rates are decomposed in three sets of rates conditional on conjugal situation. The usual TFR can be
interpreted as a weighted sum of the age- specific fertility rates of the three conjugal situations, the
weights being the proportions of women living i n each of the three conjugal situations (Laplante and
Fostik 2015).
Given that, as a rule, women do not spend all of their reproductive years in a conjugal union, and
that most children are born to women who live in a union, the TFR of women in marriage and in
consensual union are m uch higher than the TFR for all women. The usual TFR is a measure of period
fertility for all women in a given society. The TFR based on some fixed characteristic, such as ethnic group, is a measure of period fertility in the groups defined by these charac teristics. The TFR based
on a time -varying characteristic, such as conjugal situation, cannot be interpreted as an
approximation of completed fertility, as they rely on the unrealistic assumption of being continuously
married or cohabiting from age 15 to 49. However, they are measures of the intensity of fertility within these groups and are a sound and convenient way to compare fertility across such groups.
Poisson regression
The Poisson distribution is a discrete probability distribution for the counts of events that occur
randomly in a given interval of time (Evans, Hastings, and Peacock 2000). Poisson regression is a
common tool in epidemiology and demography for estimating rates and the effects of independent
variables on rates (Rodriguez and Cleland 1988, Schoumaker 2004). Poisson regression has several
advantages for studying fertility. Using it with a piecewise equation allows estimating age -specific
rates, and the sum of these rates provides the TFR.
In our calculations, we include the births that occurred in the twelve -month period that preceded
the census, i.e. the children less than 1 year old that coresided with their mothers at the time of census. Piecewise equations do not include a coefficient for the intercept; the degree of freedom

11
usually associated with the intercept is used to estimate the exact value—i.e. the ASFRs estimate s—
associated with each age.
Poisson regression allows using independent variables as any other regression model, but results
from preliminary analyses showed that in this particular case, doing so was not the best strategy ( see
details in the Results section below). We estimate the conditional ASFRs and the TFR separately for
each level of education within marriage and within cohabitation. The estimated TFRs are equal to
those that would be computed using a simple arithmetical approach, but estimating them with
Poisson regression allows us to estimate standard errors and to test the statistical significance of
observed differences. See the Technical Appendix for more de tails.
Results
We present our results in graphic form for each census and country. Figure 1 depicts age -specific
fertility rates for all women by conjugal situation, and Figure 2 represents point estimates and 95%
confidence intervals of the TFR for women in marriage and in consensual union. Detailed estimates,
statistical tests, and related statistics are reported in the Annex Table.
Conditional age -specific fertility rates
Figure 1 depicts age -specific fertility rates for women aged 15 to 49 in marriage, consensual union
and not in union. The distribution of these rates for marriage and for consensual union bears little
resemblance with the distribution of age -specific rates computed for all women, which commonly
peaks around 25. The difference stems from the fact that most women neither are married nor in a
consensual union before their 20s, but those who live with a partner at young ages are likely to have a
child soon after they start their conjugal relationship. The ASFRs for marriage and consensual un ion
reflect this pattern. As we stressed earlier, they are not an approximation of completed fertility, but
measures of the intensity of fertility within each conjugal situation.
The main result that stems from the graphs of the conditional ASFRs is that g enerally there is
great resemblance in the fertility levels of married women and women living in a consensual union across all ages. The largest differences, when there are any at all, are found among the youngest
women —aged 15 to 20 —,but these differences are not regular across countries or over time, and
could be partly due to random noise if the number of partnered women below age 20 is small. Beyond age 20, there are only minor differences in the direction of slightly higher fertility within

12
consensual union than marriage. The overall picture is, therefore, one of analogous reproductive
patterns of all women in union, regardless of union type. It is worth mentioning that the similarity of
fertility patterns between women in marital and non- marital unions cannot be considered a recent
phenomenon, since it was already manifest three or four decades ago, even in countries such as
Argentina or Chile, where the overall prevalence of cohabitation was low.
Conditional Total Fertility Rates
The main result that can be drawn from the TFR estimates by conjugal s ituation (Figure 2 and
Table 4) is that, with very few exceptions, the TFR of cohabiting women is higher than the TFR of married women and, in many cases, the difference is statistically significant. In the few cases where
the TFR is lower for cohabiting women than for married women (e.g. Ecuador 1982, Mexico 2000,
Uruguay 2006), the differences are small and statistically insignificant.
In most countries, the TFR of married women has declined more rapidly th an the TFR of women
living in consensual union over the last decades, although there are some exceptions such as Argentina and Uruguay. Consequently, the initial fertility gap between married and cohabiting
women has broadened over time, slightly in most c ountries, but considerably in others such as Brazil.
Conditional TFR for socioeconomic groups
Preliminary analyses showed that the effect of education on fertility is not proportional: highly
educated women have their children later. This did not come as a surprise, but ha d some practical
consequences. The conditional age -specific rates and the effects of the other independent variables
cannot be properly estimated without allowing for independent series of conditional ASFRs for each level of education within each conjugal situation . This can be done in a variety of ways. Two things
became clear after several different attempts. First, the most substantively interesting results are related to the education level. Second, the relatively small number of highl y educated women in
many samples does not permit using a method that simultaneously retains the property of estimating
coefficients that can be interpreted as conditional ASFR s, allows estimating different series of such
coefficients for each educational l evel within marriage and within consensual union, and allows
estimating the effects of the other socioeconomic characteristic we were interested in. We had to make a choice. We chose an approach that allows estimating coefficients that can be interpreted as
conditional ASFRs and focusing on the differences across education levels. We estimated the
conditional ASFRs and computed the TFR separately for each category of the socioeconomic

13
characteristics we consider —education, labor force status, rural/urban re sidence, and home
ownership—within marriage and within cohabitation. A s explained above, the estimated TFRs are
equal to those that would be computed using a simple arithmetical approach, but using Poisson
regression allows estimating standard errors, comp uting confidence intervals, and testing equality.
Results for education are reported in Figure 3. Detailed estimates, statistical tests, and related
statistics for all socioeconomic characteristics are available i n Table A1.
Some results are in line with what would be expected: low education, economic inactivity, and
rural residence are associated with higher fertility. Home ownership, however, does not show a
consistent association with fertility across countries and over time ; as pointed out by one referee, this
could be a consequence of landowning being dissociated from home -owning in slum areas . When we
compare fertility rates within -marriage and within -consensual union, the pattern of association
between socioeconomic characteristics and fertility does not diverge by type of union, except in the
case of education. In all countries, women with university education have the lowest fertility, regardless of type of union. However, the fertility gap between high -educated women and low –
educated women is larger among women in consensual union than among their married counterparts. This is so because while the fertility of women with less than primary education tends to be
considerably higher among women in consensual union than in marriage, the fertility o f women with
university education tends to be slightly lower among women in consensual union than in marriage.
Looking at the evolution over time of TFR by educational level within consensual union and
marriage provides additional insights. Two general fe atures emerge. First, fertility decreases over
time both within marriage and within consensual union in tandem with educational expansion.
Second, in most countries, the fertility gap between marriage and consensual union among the highly
educated diminishes over time. In many countries, over the last decades, having a child while living
in a consensual union has become less and less a distinctive feature of the lower class. Apparently, childbearing within a consensual union is becoming an increasingly frequent option among highly
educated women as well.
In fact, in the most recent census, fertility intensity is very similar within non-marital and marital
unions for highly educated women. Specifically, fertility within consensual union among women
with unive rsity education is not statistically different from fertility within marriage in seven out of 12
countries —Brazil, Chile, Colombia, Cuba, Ecuador, Mexico, and Peru—. Even in those countries
where fertility is lower within consensual union than marriage —Arg entina, Bolivia, Costa Rica and

14
Venezuela—, the magnitude of the gap is small. In sum, although consensual unions among the
middle and upper classes are a relatively recent phenomenon and still less frequent than in the lower
strata, our findings suggest t hat, in most Latin American countries, the role of cohabitation within the
family system is currently equivalent across social strata. Although during the 1980s cohabitation was rare among highly educated women and childbearing within cohabitation even rar er, nowadays
highly educated women are entering consensual unions not merely as a childless stage on the way to
marriage, but they are having children with their cohabiting partners as if they were married, in the
same fashion as their less educated counterparts.
Discussion and conclusion
Two main findings can be highlighted from the fertility measures computed by conjugal situation. The first is that, with few exceptions, the TFR of consensual union is not only close to, but also
slightly higher than the T FR of marriage. The second is that the distributions of ASFRs for marriage
and consensual union are very similar across the age range, except before age 20. The general conclusion we may draw from those findings is that in Latin America, at least for the last four
decades, women’s fertility patterns depend on their age and on their living in a conjugal union, but not on the legal nature of this union. For the childbearing behavior of women residing with a partner,
it does not seem to matter whether they are legally married or not.
Prior studies had already documented that childbearing within consensual unions was
widespread in some Latin American countries ( Castro -Martín 2002). In this study, we have
confirmed this pattern for 13 countries —which represent about 87% of the total population in the
region —over the last four decades. Certainly, this pattern is not unique to Latin America —the
probability of having a child is practically the same for cohabiting and married couples in several
European countries (Tou lemon and Testa 2005; Hoem and Mureșan, 2013) —, but we have
documented that it is less of a novelty in the Latin America region. The decoupling of marriage and childbearing has long taken place without a dissociation of partnership and parenthood.
However , the important contribution of this study has been to show that the similarities in
reproductive behavior between marital and non-marital unions are not confined to the socially
disadvantaged groups, but apply as well to the better off. The negative educa tional gradient of
childbearing within consensual union continues to be the predominant pattern in Latin America, as it
is in the United States and in some European countries —mostly in Central and Eastern Europe —

15
(Perelli -Harris et al., 2010; Goldstein and Kenney 2012). However, this gradient has become less
steep over time in Latin America, partly due to the increasing convergence of marital and non-marital
unions regarding childbearing behavior among the highly educated. This was not the usual pattern in
the past. Around 1980, the proportion of births to women with tertiary education that took place
within a consensual union was below 5% in all countries except Panama. Nowadays, not only are
university -educated women much more likely to enter a consensual u nion than three or four decades
ago, but their childbearing patterns do not differ much from their married counterparts .
As stated in the introduction, there is no broad consensus on whether the recent demographic
changes in Latin America should be attrib uted to modernity and the advance of the Second
Demographic Transition in the region, or to the increasing uncertainty that both lower and middle
classes are now confronted with. On one hand, recent changes in family dynamics have occurred
under far better economic conditions than in the 1980s. On the other hand, the remarkable expansion
of education—gross enrolment rates in third -level education rose from 2 3 to 42 percent between
2000 and 2011 (ECLAC 2013) —entails that some of the privilege s conferred by tertiary education in
the labor market four decades ago might have dwindled. There are some demographic changes that
point in the direction of the STD, such as the increasing number of countries approaching below
replacement fertility (Cavenaghi and Alves 2 009), the recent boom of cohabitation involving all
social strata ( Esteve, Lesthaeghe, and López -Gay 2012), the surpassing of children born inside
marriage by children born outside marriage since the turn of the century ( Castro -Martín et al. 2011 ),
and the emergence of a postponement pattern of first births —though still timid —among the higher
educated (Rosero -Bixby , Castro -Martín and Martín -García 2009). Some of the ideational features of
the STD, such as increasing secularization and tolerance to various t ypes of non -conformist
behaviors like divorce, homosexuality or euthanasia , have also been observed in most Latin
American societies ( Esteve , Lesthaeghe , and López -Gay 2012). Rising education among younger
cohorts is probably a fundamental factor driving i deational change and creating a context of growing
tolerance for different life styles and family behaviors . However, a recent study by Covre -Sussai
(2013) found that, although national contexts of higher tolerance were related to the occurrence of
cohabit ation in the higher educated groups, at the individual level, women in the upper strata were
not necessarily more tolerant in their values than lower educated ones.
As it is the case in more developed countries (Heuveline and Timberlake 2004, Sassler and
Miller 2011; Hiekel and Castro -Martín 2014), the meaning attached to cohabitation—as well as the

16
perceived advantages of marriage—is likely to differ across societies in Latin America and across
women with different social background (Covre -Sussai et al. 2 014). Our a priori expectations, in line
with traditional patterns and what is still the most common opinion, were that consensual union in the
lower social strata would be related to economic constraints and would serve as a surrogate for
marriage , and that consensual union in the upper social strata would be related to women’s
empowerment and would serve as an adaptative strategy to postpone motherhood until marriage.
However, the findings from this study did not fully confirm these expectations. Whereas childbearing
within consensual union among highly educated women was uncommon in the 1980s, nowadays
fertility levels within cohabitation and marriage are alike among the higher educated. Therefore, the
emergent pattern of cohabitation among the upper stra ta cannot be merely considered as a transitory
prelude to marriage, but a much longer life cycle stage where childbearing takes place, as it has
traditionally happened in the lower strata.
In sum, in addition to the coexistence of marriage and consensual union in the Latin American
region, we should be aware of the coexistence of several types of consensual unions: some are linked
to historical cultural legacies, some are poverty -driven or a strategy to cope with unplanned
pregnancy, and some favor interpe rsonal commitment above institutional regulation. Consensual
unions in the higher and lower strata probably have different motives and rationale s, and
acknowledging the existing diversity of consensual unions and the corresponding consequences in
the event of dissolution for the well -being of children is still an unsettled challenge for public
policies. The adoption of cohabitation by the highly educated has probably been facilitated by the
wide social recognition conferred to consensual unions in the lower social strata, as this behavior did
not run counter to existing moral and legal codes. Likewise, it is possible that the diffusion of
childbearing within consensual union in the upper social strata may have been facilitated by changes in other aspects of family life. Among those, two maybe of special importance: the increase in the
labor force participation of women and the advances in the economic protection of children in the
event of union disruption that have been implemented in most Latin American cou ntries. Some other
institutional factors may be at play. Most Latin American countries typically maintain a somewhat decent system of public health: as in most European countries, access to basic health care is universal
and unrelated with family status. S uch a factor could be among the reasons why family formation
within unmarried cohabitation is increasing faster in Latin America and, for instance, in the UK, than

17
in the USA . Future research on the diffusion of childbearing within consensual union in Lati n
America could examine such hypotheses.
Technical Appendix
The own -children method of fertility estimation
The own- children method is an indirect technique for the estimation of fertility by age using census
data (Cho, Rutherford, and Choe 1986). Its orig inal form uses the distribution of the number of
children less than five years old in the household conditional on the age of mothers aged between 15
and 49, grouped into five -year classes. It was developed for the USA census, mainly to relate fertility
measures with characteristics available in the census, but not in vital statistics. Using the number of
children less than five years old allowed comparability with the “fertility ratio”, a measure based on
this number and commonly used at the time the origi nal authors introduced the method. The most
obvious difficulties and limitations of this method are establishing the relationship between mother and child from census records, census undercoverage of children and women, infant mortality, and
children who do not live with their mother (Grabill and Cho 1965).
In our analyses, we use the information provided in IPUMS files on the age of the youngest child
of women aged between 15 and 49 (cf. Sobek and Kennedy 2009). We use only the births that
occurred in the year preceding the census, i.e. children less than 1 year of age at census date. This
would not be an optimal strategy if our main objective were to estimate age -specific rates with small
variance, but it is better suited to our goal of comparing marital and non-marital unions than using the
original five -year period. Some women may have got married between the birth of the child and the
time of census, and some women who were not living with a partner may have started living in a consensual union after the birth of the child. In addition, some women married or living in a
consensual union at the time the child was born may be living alone at the time of census. These
potential changes in union status are less likely over a one -year period than over a five -year period.
Nevertheless, if the birth of a child tends to trigger the transformation of a consensual union into marriage in the first year of the child’s life, the within -marriage ASFR s and TFR will be
overestimated whereas the within -consensual union AS FRs and TFR will be underestimated.

18
Poisson regression
Formally, for any given individual, the Poisson regression model we use may be written as
3 49
1 15ˆij
ij i j
ijr st eα
= ==∑∑
with
() ,!ij ijry
ij
ij
ijerfyy−
=
where ˆijr is the predic ted conditional age -specific fertility rate for age t j within conjugal situation s i, si
and tj are binary indicator variables for conjugal situation i and age j respectively, αij is the element
of the vector of coefficients for conjugal situation i and age j, yij is the total number of births to
women in conjugal situation i and of age j, f(yij) is the density function and yij ! stands for the factorial
of y ij, i.e. the product of all integers from y ij through 1.
Given that by design, the coefficients are n ot correlated, the variance of the sum of the
conditional age -specific rates is the sum of their variances. Formally,
49
15var( ) var( ),i ij
jRr
==∑
where R i is the conditional TFR for conjugal situation i .
References
Abramo, Laís and María Elena Valenzuela. 2005. “Women’s labour force participation rates in Latin
America,” International Labour Review 144( 4): 369–399.
Arriagada, Irma. 2002. “ Changes and inequa lity in Latin American families,” Cepal Review 77: 135–
153.
———. 2007. “Familias latinoamericanas: C ambiantes, diversas, desiguales,” Papeles de Población
53: 9–22.
Billari, Francesco and Aart Liefbroer. 2004. “Is the Second Demographic Transition a useful concept
for demography? Introduction to a debate,” Vienna Yearbook of Population Research 2004: 1–3.
http://hw.oeaw.ac.at/0xc1aa500d_0x0006201c.

19
Binstock, Georgina. 2010. “ Tendencias sobre la convivencia, matrimonio y maternidad en áreas
urbanas de Argentina,” Revista Latinoamericana de Población 3(6): 129–146.
Brown, Elizabeth and Alfred Dittgen. 2000. “Fertility of married and unmarried couples in Europe,”
paper presented at the FFS flagship conference Partnership and fertility -A revolution?, Brussels,
29–31 May 2000.
Cabella, Wanda. 2009. “Dos décadas de transformaciones de la nupcialidad uruguaya,” Estudios
Demográficos y Urbanos 24(2): 389–427.
Cabella, Wanda, Andrés Peri, and Constanza Street. 2005. “¿Dos orillas y una transición? La segunda
transición demográfica en Buenos Aires y Montevideo en perspectiva biográfica,” in Susanna
Torrado (ed.), T rayectorias Nupciales, familias ocultas (Buenos Aires, entresiglos) . Buenos
Aires: Miño y Dávila, pp. 207–231.
Cavenaghi, Suzana, and José Eustáquio Alves. 2009. “Fertility and Contraception in Latin América:
Historical Trends, Recent Patterns.” In Demographic Transformations and Inequalities in Latin America: Historical Trends and Recent Patterns , edited by S. Cavenaghi, Chapter 6, 161–193.
Serie Investigaciones 8. Rio de Janeiro: Latin American Population Association.
Castro -Martín , Teresa. 1995. “Women’ s education and fertility: Results from 26 Demographic and
Health Surveys,” Studies in Family Planning 26(4): 187–202.
———. 2001. “Matrimonios sin papeles en Centroamérica: persistencia de un sistema dual de
nupcialidad,” in Luis Rosero- Bixby (ed.), Poblac ión del Istmo 2000: Familia, migración,
violencia y medio ambiente . San José, Costa Rica, Centro Centroamericano de Población,
pp. 41–65. http://ccp.ucr.ac.cr/libros/poblaist/pdf/poblacion_istmo.pdf
———. 2002. “Consensual unions in Latin America: persistence of a dual nuptiality system,”
Journal of Comparative Family Studies 33(1): 35–55.
Castro -Martín , Teresa, Clara Cortina, Teresa Martín -García and Ignacio Pardo. 2011. “Maternidad
sin matrimonio en América Latina: un análisis comparativo a partir de datos censales,” Notas de
Población 93: 37–76.

20
Castro -Martín , Teresa, Teresa Martín -García and Dolores Puga González. 2008. “Tipo de unión y
violencia de género: una comparación de matrimonios y uniones consensuales,” in Laura
Rodríguez Wong (org.), Población y salud sexual y reproductiva en América Latina. Río de
Janeiro, Asociación Latinoamericana de Población.
http://www.alapop.org/docs/publicaciones/investigaciones/SSR_parteIV -1.pdf
CEPAL. 2002. Vulnerabilidad sociodemográfica: viejos y nuevos riesgos para c omunidades, hogares
y personas. Brasilia, 6 -10 Mayo 2002. LC/R.2086.
http://www.eclac.cl/publicaciones/xml/4/10264/LCR2086_ind.pdf
Cho, Lee -Jay, Robert D. Rutherford, and Minja Kim Choe. 1986. The own -children method of
fertility estimation . Honolulu HI: U niversity of Hawaii Press.
Covre -Sussai, Maira . 2013. Cohabitation in Latin America and developed countries: A cross -national
perspective. International Journal of Humanities and Social Science 3(16): 29- 43.
Covre -Sussai, Maira, Jan Van Bavel, Koenraad Mat thijs and Gray Swicegood. 2014. Disentangling
the Different Types of Cohabitation in Latin America: Gender Symmetry and Contextual Influences.
Available at SSRN: http://ssrn.com/abstract=2376739 or http://dx.doi.org/10.2139/ssrn.2376739 .
De V os, Susan. 2000. “Nuptiality in Latin America,” in Sandra Lee Browning and R. Robin Miller
(eds.), Till Death Do Us Part: A Multicultural Anthology on Marriage . Stamford, CT: JAI Press,
pp. 219–243.
Do Valle Silva, Nelson, Maria Helena F. T. Henriques and Amaury de Souz a. 1990. An analysis of
reproductive behavior in Brazil , Instituto de Estudos Econômicos, Sociais e Politicos de São
Paulo and Population Council (Demographic and Health Surveys Further Analysis Series nș 6)
Dumas, Jean and Alain Bélanger. 1998. Report on the Demographic Situation in Canada 1996 .
Ottawa, Demography Division, Statistics Canada.
Economic Commission for Latin America and the Caribbean (ECLAC). 2013. Statistical Yearbook
for Latin America and the Caribbean 2013. Santiago de Chile: United Nations .
Esteve, Albert, Ron Lesthaeghe, and Antonio López -Gay. 2012. “The Latin American cohabitation
boom, 1970- 2007,” Population and Development Review 38(1): 55–81
Evans, Merran, Nicholas Hastings, and Brian Peacok. 2000. Statistical distributions, 3
rd edition. New
York, NY: John Wiley and Sons.

21
García, Brígida and Olga Rojas. 2004. “Las uniones conyugales en América Latina:
Transformaciones en un marco de desigualdad social y de género,” Notas de Población 78: 65–
96.
Goldstein, Joshua R. and Catherine T. Kenney. 2012. “Too poor to marry? A cross -national
comparison of the SES gradient in non- marriage ”, Paper presented at the European Population
Conference, Stockholm, 13- 16 June 2012.
Grabill, Wilson H. and Lee -Jay. Cho. 1965. “Methodology for the measurement of current fertility
from population data on young children,” Demography 2: 50–73.
Greene, Margaret E. 1991. The importance of being married: Marriage choice and its consequences
in Brazil . Doctoral Dissertation, University of Pennsylvania.
Harbitz, Mia Elisabeth, Juan Carlos Benítez Molina, and Iván Arcos Axt . 2010. Inventario de los
registros civiles e identificación de América Latina y el Caribe. Washington DC: Banco
Interamericano de Desarrollo.
Heuveline, Patrick and Jeffrey M. Timberlake. 2004. “The r ole of cohabitation in family formation:
The United States in comparative perspective,” Journal of Marriage and Family 66(5): 1214–
1230.
Hiekel, Nicole and Teresa Castro -Martín. 2014. “ Grasping the diversity of cohabitation: Fertility
intentions among coha biters across Europe,” Journal of Marriage and Family 76(3): 489–505.
Hoem, Jan M., Cornelia Mureșan and Mihaela Hărăguș. 2013. Recent features of cohabitational and
marital fertility in Romania. Max Planck Institute f or Demographic Research, Working Paper
WP 2013- 007.
Hoem, Jan M. and Cornelia Mureșan. 2011. “The Total Marital Fertility Rate and its extensions,”
European Journal of Population 27:295–312
Kiernan, Kathleen. 2001. “The rise of cohabitation and childbearing outside marriage in Western
Europe,” International Journal of Law, Policy and the Family 15(1): 1–21.
Kiernan, Kathleen, Sara McLanahan, John Holmes and Melanie Wright. 2011. Fragile Families in the
US and UK. Center for Research on Child Well -Being, Princeton University, Working Paper WP
11-04-FF.
Klüsener, Sebastian , Brienna Perelli -Harris , and Nora Sánchez Gassen . 2013. “Spatial aspects of the
rise of non- marital fertility across Europe since 1960: The role of states and regions in
shaping patterns of change,” European Journal of Populati on 29:137–165.

22
Laplante, Benoît and Ana Laura Fostik. 2015. “Two period measures for comparing the fertility of
marriage and cohabitation,” Demographic Research 32(14): 421–442.
Laplante, Benoît and María Constanza Street. 2009. “Los tipos de unión consensual en Argentina
entre 1995 y 2003: una aproximación biográfica,” Estudios Demográficos y Urbanos
24(2): 351–87.
Lesthaeghe, Ron. 1995. “The Second Demographic Tra nsition in Western countries: An
interpretation,”.in Karen Oppenheim Mason and An- Magritt Jensen (eds.), Gender and Family
Change in Industrialized Countries . Oxford, UK: Clarendon Press, pp. 17– 62.
———. 2010. “The unfolding story of the Second Demographic Transition,” Population and
Development Review 36(2): 211–252.
Minnesota Population Center. 2013. Integrated public use microdata series, international: Version
6.2 [Machine- readable database]. Minneapolis, MN: University of Minnesota.
Parrado, Emilio and Marta Tienda. 1997. “Women’s roles and family formation in Venezuela: New
forms of consensual unions?,” Social Biology 44(1- 2): 1–24.
Perelli -Harris, B., Sigle -Rushton, W., Kreyenfeld, M., Lappegård, T., K eizer, R., Berghammer, C.
Perelli -Harris, Brienna, Wendy Sigle -Rushton, Michaela Kreyenfeld, Trude Lappegård, Renske
Keizer, and Caroline Berghammer. 2010. “The educational gradient of childbearing within
cohabitation in Europe”, Population and Development Review 36(4): 775- 801.
Quilodrán, Julieta. 1999. “L’union libre en Amérique Latine: aspects récents d’un phénomène
séculaire,” Cahiers Québécois de Démographie 28(1 -2): 53–80.
———. 2011. “Un modelo de nupcialidad post -transicional en América Latina?” In G. Binstock and
J. Melo (Eds.), Nupcialidad y famil ia en la América Latina actual (pp. 11–33). Rio de Janeiro:
ALAP.
Quilodrán, Julieta and Teresa Castro -Martín . 2009. “Nuevas dinámicas familiares,”. Estudios
Demográficos y Urbanos 24(2): 283–912.
Raley, R. Kelly. 2001. “Increasing fertility in cohabiting unions: Evidence for the Second
Demographic Transition in the United States?,” Demography 38(1): 59–66.
Rodriguez, German and John Cleland. 1988. “Modelling marital fertility by age and duration: An
empirical appraisal of the Page model,” Population Studie s 42(2): 241–257.

23
Rodríguez , Jorge. 2005. “Cohabitación y matrimonio: Disparidades en el origen y en el desempeño y
sus implicaciones en materia de protección social,” in Irma Arriagada (ed.), Políticas hacia las
Familias, Protección e Inclusión Social. Sa ntiago: CEPAL (Serie Seminarios y Conferencias).
Rodríguez Vignoli, Jorge. 2004. “Cohabitación en América Latina: ¿Modernidad, exclusión o
diversidad?,” Papeles de Población 40: 97–145.
———. 2011. “La situación conyugal en los censos latinoamericanos de la década de 2000:
relevancia y perspectivas,” in Magda Ruiz Salguero and Jorge Rodríguez Vignoli (eds.), Familia
y nupcialidad en los censos latinoamericanos recientes: una realidad que desborda los datos .
Santiago: CELADE (Serie Población y Desarrollo nș 99), pp. 47–70.
http://www.eclac.org/publicaciones/xml/9/42709/lcl3293e -P.pdf
Rosero -Bixby, Luis, Teresa Castro -Martín and Teresa Martín -García. 2009. “Is Latin America
starting to retreat from early and universal childbearing?” Demographic Research, V olum e 20,
Article 9, pp. 169- 194.
Sassler, Sharon and Amanda J. Miller. 2011. Class differences in cohabitation processes. Family
Relations 60(2): 163–177.
Schoumaker, Bruno. 2004. “Une approche personnes -périodes pour l’analyse des histoires
génésiques,” Population 59(5): 783–796.
Seltzer, Judith A. 2000. “Families formed outside of marriage,” Journal of Marriage and the Family
62(4): 1247–1268.
Sobek, Matthew and Sheela Kennedy. 2009. The development of family interrelationship variables
for international census data. Minnesota Population Center (Working Paper No. 2009–02)
Sobotka, Thomas and Laurent Toulemon. 2008. “Changing family and partnership behaviour:
Common trends and persistent diversity across Europe,” Demographic Research, vol. 19, art. 6
Socolow, Susan M. 2000. The Women of Colonial Latin America. NY: Cambridge University Press.
Toulemon, Laurent and Maria Rota Testa. 2005. Fertility intentions and actual fertility: A complex
relationship. Population and Societies no. 415.
Van Hook, Jennifer and Cl aire E. Altman. 2013. “Using Discrete -Time Event History Fertility
Models to Simulate Total Fertility Rates and Other Fertility Measures,” Population Research
and Policy Review 32:585–610.

24
Verdugo Lazo, Aida. 1994. Marital fertility in Brazil: Differential by type of union and its importance
in the fertility transition, 1976 -1991. Calverton MD: Macro International Inc. (DHS Working
Papers nș 15).

25
TABLE 1 Proportion of women living in consensual union among
women aged 15- 49 in conjugal union, according to most recent data
source
Percentage Source and date
Dominican Republic 73.9 DHS 2007
Panama 64.1 Census 2010
Honduras 62.3 DHS 2011 -2012
Nicaragua 59.8 RHS 2006/07
Peru 59.5 DHS 2012
Colombia 58.3 DHS 2010
Uruguay 52.7 Census 2011
Cuba 49.4 Census 2002
El Salvador 48.9 Census 2007
Venezuela 47.8 Census 2001
Ecuador 43.8 RHS 2004
Brazil 43.5 Census 2010
Costa Rica 39.6 Census 20 11
Paraguay 39.5 RHS 2004
Bolivia 37.0 DHS 2008
Guatemala 33.1 RHS 2002
Mexico 33.0 Census 2010
Argentina 30.6 Census 2001
Chile 19.8 Census 2002
Sources: Census, Demographic and Health Surveys (DHS); Reproductive Health
Surveys (RHS).

26

TABLE 2 Proportion of births from unmarried mothers according to vital statistics
Year Percentage Year Percentage
Argentina 1980 29.8 2000 57.6
Chile 1970 18.8 2010 68.0
Costa Rica 1970 29.4 2010 67.4
Cuba — — 2011 80.7
Ecuador 1966 32.0 — —
El Salvador 1970 67.8 1998 72.8
Guatemala 1970 61.9 — —
Mexico 1970 27.3 2010 59.4
Pana ma 1970 70.9 2002 79.9
Paraguay 1970 42.6 2011 67.8
Peru 1972 41.3 2000 58.3
Dominican Republic — — 2011 89.8
Uruguay 1970 21.1 2001 55.2
Venezuela 1970 38.8 2010 83.5
Sources: United Nations Demographic Yearbook; World Fertility Report; National Institutes of
Statistics

27

TABLE 3 Evolution of the distribution of births according to the conjugal
situation of the mother in selected countries from Latin America, 1970 -2000
1970 1980 1990 2000
Not in umion 7.3 7.7 10.9 15.0
Consensual union 16.8 18.0 26.8 38.9
Marriage 75.9 74.3 62.3 46.1
Source: Census microdata, IPUMS -International. Countries included: Argentina , Bolivia,
Brazil, Chile, Colombia, Costa Rica, Cuba, Ecuador, Mexico, Panama, Peru, Uruguay, and
Venezuela.

28
TABLE 4 Estimates of the total fertility rate of women living in sele cted Latin American countries according to their conjugal
situation using data from censuses of the four most recent censuses round. Estimates from the own -children method. Women
aged 15 -49. Census data from IPUMS. Weighted estimation.
Argentina Bolivia Brazil Chile
1980 1991 2001 2001 1980 1991 2000 2010 1982 1992 2002
TFR
All women 2.90 2.71 2.30 2.84 3.67 2.40 2.02 1.57 2.36 2.21 1.58
Married 5.34 4.56 4.65 4.33 6.27 4.38 3.52 2.64 4.69 3.79 2.93
In consensual union 5.90*** 5.73*** 4.74 5.15*** 6.54*** 4.93*** 4.28*** 3.18*** 4.72 4.22*** 3.22**
Alone 0.59 0.72 0.77 1.01 0.64 0.68 0.78 0.64 0.80 0.87 0.79
Unknown situation — — — — 2.19 3.52 — — — — —
Colombia Costa Rica Cuba Ecuador Mexico
1985 1993 2005 1984 2000 2002 1982 1990 2001 2010 2000 2010
TFR
All women 2.27 1.88 1.90 3.02 2.23 1.51 3.72 2.74 2.11 1.98 2.30 1.96
Married 3.92 3.09 3.20 4.86 3.80 2.23 5.77 4.51 3.39 3.35 4.29 3.74
In consensual union 4.44*** 3.55*** 3.72*** 5.34 4.35** 2.42* 5.71 4.72* 3.53 3.45 4.23 3.86
Alone 0.74 0.65 0.88 1.35 1.06 0.58 1.07 0.71 0.67 0.83 0.60 0.66
Unknown situation 0.63 0.42 0.67 — — — 1.18 0.75 0.80 — 1.24 0.43
Panama Peru Uruguay Venezuela
1980 1990 2000 2010 1993 2007 1985 1996 2006 1981 1990 2001
TFR
All women 3.05 2.50 2.60 2.12 2.89 1.98 2.38 2.24 1.68 3.23 2.77 1.93
Married 4.44 3.99 3.88 3.10 4.72 3.24 4.73 3.90 3.79 5.07 4.58 3.27
In consensual union 5.25*** 4.53* 4.66* 3.94** 5.49*** 3.69*** 4.98 4.70** 3.56 6.17*** 5.41*** 3.88***
Alone 1.27 1.02 1.15 0.92 0.79 0.64 0.56 0.76 0.64 1.30 1.21 0.78
Unknown situation — — — — 0.98 — — — — 1.07 0.32 0.44
*: p<0.05; **: p<0.01; ***: p<0.001. Each test compares the estimated TFR for consensual union to the corresponding estimat e for
marriage.

29
TABLE A1 Estimates of the total fertility rate of women living in sele cted Latin American countries according to selected
socioeconomic characteristics by conjugal situation using data from censuses of the four most recent censuses round. Estimate s
from the own -children method and Poisson regression. Women aged 15 -49. Census data from IPUMS. Weighted estimation.
Argentina Chile
Marriage Consensual union Marriage Consensual union
1980 1991 2001 1980 1991 2001 1982 1992 2002 1982 1992 2002
Education level
Less than primary 5.67 4.64 5.19 6.68*** 6.49*** 5.92* 4.71 3.37 2.65 5.23 4.46*** 3.17
Primary 5.20 4.53 4.45 5.12 5.54*** 4.73* 4.57 3.70 2.84 4.53 4.23*** 3.19**
Secondary 3.94 4.21 4.03 4.13 3.43** 3.91 3.98 3.59 2.63 3.25 3.83 3.07
University 4.40 3.32 3.48 3.58 3.85** 2.20*** 3.09 3.26 2.42 1.50*** 3.03 1.85
Labor force status
Employed 3.48 3.43 3.22 2.60 4.01*** 3.49 2.97 2.64 1.95 2.25 1.99** 1.76
Unemployed 5.57 3.45 3.82 10.37*** 4.35** 4.13 3.38 2.61 1.47 2.87 1.03** 1.91
Inactive 5.79 5.16 5.32 6.73*** 6.81*** 5.70** 4.93 3.99 3.33 5.17 4.57*** 3.79***
Area
Rural 6.19 5.11 5.00 7.45*** 6.81*** 5.84*** 5.29 4.04 3.00 5.36 5.28*** 3.42
Urban 5.25 4.53 4.58 5.50 5.60*** 4.57 4.59 3.74 2.90 4.59 4.03* 3.19*
Home ownership
Owner 5.05 4.27 4.51 5.43* 5.60*** 4.68 4.53 3.65 2.87 4.47 3.99* 3.15*
Other 5.96 5.18 4.95 6.48** 5.99*** 4.97 4.99 4.10 3.09 5.11 4.64* 3.37
*: p<0.05; **: p<0.01; ***: p<0.001. Each test compares the estimated TFR for consensual union to the corresponding estimate for
marriage.

30

TABLE A1 Estimates of the total fertility rate of women living in selected Latin American countries according to selected
socioeconomic characteristics by conjugal situation using data from censuses of the four most recent censuses round. Estima tes
from the own -children method and Poisson regression. Women aged 15 -49. Census data from IPUMS. Weighted estimation.
(Continued).
Bolivia Brazil
Marriage Consensual Marriage Consensual union
2001 2001 1980 1991 2000 2010 1980 1991 2000 2010
Educa tion level
Less than primary 5.22 5.77* 6.50 4.64 3.83 2.56 6.77*** 5.19*** 4.55*** 3.37***
Primary 4.04 4.68*** 5.33 3.92 3.23 2.54 5.14 4.38*** 3.97*** 3.20***
Secondary 3.36 4.05* 4.75 3.73 3.08 2.54 4.01*** 3.87 3.54*** 2.87*
Universi ty 2.74 2.44*** 3.84 2.82 2.37 2.19 2.93*** 3.13 2.34 1.94
Labor force status
Employed 3.67 4.08 4.47 3.19 2.55 1.86 4.58 3.29 2.85*** 2.06***
Unemployed 2.78 4.37*** 5.81 3.68 2.56 1.97 6.24 4.22* 3.46*** 2.58***
Inactive 4.79 5.93*** 6.66 4.80 4.32 3.63 7.12*** 5.55*** 5.30*** 4.26***
Area
Rural 5.16 5.89** 7.39 5.28 4.12 2.67 7.76*** 6.02*** 5.04*** 3.48***
Urban 3.88 4.81*** 5.78 4.10 3.36 2.62 6.12*** 4.66*** 4.12*** 3.11***
Home ownership
Owner 4.34 5.02*** 6.30 4.39 3.46 2.57 6.80*** 4.98*** 4.21*** 3.12***
Other 4.41 5.41*** 6.28 4.41 3.66 2.77 6.35 4.90*** 4.38*** 3.28***
*: p<0.05; **: p<0.01; ***: p<0.001. Each test compares the estimated TFR for consensual union to the corresponding estimate for
marriage.

31

TABLE A1 Estimates of the total fertility rate of women living in selected Latin American countries according to selected
socioeconomic characteristics by conjugal situation using data from censuses of the four most recent censuses round. Estimat es
from the own -children method and Poisson regression. Women aged 15 -49. Census data from IPUMS. Weighted estimation.
(Continued).
Colombia Costa Rica
Marriage Consensual union Marriage Consensual union
1985 1993 2005 1985 1993 2005 1984 2000 1984 2000
Education level
Less than primary 4.34 3.33 3.57 4.66** 3.88*** 4.18** 5.07 3.73 5.75 4.53*
Primary 3.68 2.96 2.96 4.16*** 3.34*** 3.54*** 4.77 3.78 4.80 4.45**
Secondary 3.16 2.95 2.91 3.97** 3.08 3.31 3.50 3.42 3.85 2.27**
Universi ty 2.73 1.83 2.40 3.20 1.79 2.64 4.59 2.70 0.33*** 1.41**
Labor force status
Employed 3.36 1.91 2.14 3.69** 2.34*** 2.28 2.97 2.12 2.76 1.92
Unemployed 2.49 2.03 2.87 3.07 1.90 2.98 4.86 1.92 3.25** 1.83
Inactive 4.20 3.52 3.65 4.82*** 4.01*** 4.20*** 5.10 4.19 5.59 4.90**
Area
Rural 4.54 3.81 3.44 4.65 4.35*** 4.06*** 5.23 3.87 5.60 4.63**
Urban 3.63 2.85 3.11 4.37*** 3.22*** 3.61** 4.58 3.80 4.76 4.06
Home ownership
Owner 3.69 2.90 3.06 4.22*** 3.44*** 3.66*** 4.85 3.56 5.45 4.22*
Other 4.31 3.33 3.55 4.80*** 3.72*** 3.91 4.90 4.22 5.36 4.47
*: p<0.05; **: p<0.01; ***: p<0.001. Each test compares the estimated TFR for consensual union to the corresponding estimate for
marriage.

32

TABLE A1 Estimates of the to tal fertility rate of women living in selected Latin American countries according to selected
socioeconomic characteristics by conjugal situation using data from censuses of the four most recent censuses round. Estimate s
from the own -children method and Po isson regression. Women aged 15- 49. Census data from IPUMS. Weighted estimation.
(Continued).
Cuba Ecuador
Marriage Consensual Marriage Consensual union
2002 2002 1982 1990 2001 2010 1982 1990 2001 2010
Education level
Less than primary 2.03 2.47 6.81 5.20 3.75 3.77 6.47 5.21 3.91 3.68
Primary 2.16 2.34 4.96 4.26 3.30 3.22 4.82 4.46 3.45 3.46**
Secondary 2.00 2.12 3.93 3.76 2.89 2.90 3.21 4.61* 2.65 3.36**
University 1.65 1.82 3.80 2.73 2.15 2.37 0.45*** 2.20 1.93 2.27
Labor force status
Employed 1.52 1.93 4.25 3.42 2.66 2.75 3.63 3.30 2.42 2.62
Unemployed 1.47 0.93 2.75 4.60 1.20 2.76 6.17*** 3.49 2.62** 3.04
Inactive 2.52 2.61 6.07 4.91 3.83 3.83 5.92 5.01 3.91 3.84
Area
Rural — — — 5.40 3.87 3.72 — 5.23 4.14 3.79
Urban — — — 4.13 3.21 3.09 — 4.51** 3.33 3.24
Home ownership
Owner — — 5.74 4.39 3.37 3.23 5.67 4.64* 3.39 3.31
Other — — 5.77 4.68 3.43 3.55 5.88 4.88 3.79** 3.66
*: p<0.05; **: p<0.01; ***: p<0.001. Each test compares the estimated TFR for consensual union to the corresponding estimate for
marriage.

33

TABLE A1 Estimates of the total fertility rate of women living in selected Latin American countries according to selected
socioeconomic characteristics by conjugal situation using data from censuses of the four most recent censuses round. Estimates
from the own -children method and Poisson regression. Women aged 15 -49. Census data from IPUMS. Weighted estimation.
(Continued).
Mexico Panama
Marriage Consens ual Marriage Consensual union
2000 2010 2000 2010 1980 1990 2000 2010 1980 1990 2000 2010
Education level
Less than primary 4.62 3.70 4.44 4.29*** 5.98 4.46 6.09 2.83 5.29 4.84 5.52 5.05***
Primary 4.06 3.61 4.05 3.72 3.92 4.07 3.20 3.59 5.06*** 4.48 4.59*** 3.95
Secondary 3.66 3.23 3.52 3.49 3.40 3.94 3.19 2.41 3.75 3.26 4.50*** 2.63
University 3.07 2.61 2.22*** 2.59 2.38 2.06 2.00 1.81 2.20 3.78** 3.87*** 2.01
Labor force status
Employed 3.00 2.59 2.83 2.76 2.94 2.66 2.49 2.03 3.62 3.36 3.01 2.55
Unemployed 2.35 2.40 1.66 1.93 1.96 2.13 3.62 2.28 3.08 3.63* 3.98 2.51
Inactive 4.67 4.09 4.70 4.28** 4.92 4.18 4.49 3.61 5.77** 4.80* 5.22 4.61**
Area
Rural 4.81 — 4.74 — 5.33 — 5.44 3.39 5.65 — 5.17 4.57*
Urban 4.13 — 4.07 — 3.76 — 3.20 2.99 4.65** — 4.27*** 3.39
Home ownership
Owner 4.22 3.68 4.14 3.85* 4.59 4.07 4.86 2.85 5.88*** 4.55 4.63 4.02***
Other 4.50 3.92 4.43 3.91 4.20 3.95 3.07 3.38 4.59 4.53 4.83*** 3.63
*: p<0.05; **: p<0. 01; ***: p<0.001. Each test compares the estimated TFR for consensual union to the corresponding estimate for
marriage.

34

TABLE A1 Estimates of the total fertility rate of women living in selected Latin American countries according to selected
socioeconom ic characteristics by conjugal situation using data from censuses of the four most recent censuses round. Estimates
from the own -children method and Poisson regression. Women aged 15 -49. Census data from IPUMS. Weighted estimation.
(Continued).
Peru Urug uay
Marriage Consensual union Marriage Consensual union
1993 2007 1993 2007 1985 1996 2006 1985 1996 2006
Education level
Less than primary 5.41 3.48 5.86*** 4.24*** 4.24 3.34 † 5.97** 5.23*** 2.84
Primary 4.32 3.12 5.23*** 3.68*** 4.72 3.90 † 4.61 4.75** 3.70
Secondary 4.37 2.97 4.60 3.12 5.31 3.92 † 3.68* 1.84*** 3.18
University 3.12 2.54 2.88 2.23 3.09 2.36 † 1.00*** 0.75*** 1.31
Labor force status
Employed 3.66 2.18 4.37*** 2.63*** 2.97 3.48 1.96 3.55 2.57* 2.22
Unemployed 2.89 2.95 4.39** 2.67 2.32 2.53 2.27 2.83 4.79*** 2.54
Inactive 5.12 3.71 5.83*** 4.20*** 5.33 4.78 4.68 5.92 6.54*** 4.96
Area
Rural 5.96 3.70 6.28* 4.44*** 4.60 — — 5.79 — —
Urban 4.31 3.05 5.11*** 3.40*** 4.71 — — 4.92 — —
Home ownership
Owner 4.75 3.17 5.46*** 3.70*** 4.62 3.95 3.29 5.12 4.64 3.38
Other 4.71 3.46 5.62*** 3.72 4.76 3.96 3.66 5.03 4.77* 3.86
*: p<0.05; **: p<0.01; ***: p<0.001. Each test compares the estimated TFR for consensual union to the corres ponding estimate for
marriage. †The 2006 “census data” from Uruguay actually come from a survey. The small size of the subsample of married women by
education level leads to inconsistent estimates of the TFR.

35

TABLE A1 Estimates of the total fertility ra te of women living in selected Latin American countries according to selected
socioeconomic characteristics by conjugal situation using data from censuses of the four most recent censuses round. Estimate s
from the own -children method and Poisson regression. Women aged 15- 49. Census data from IPUMS. Weighted estimation.
(Continued).
Venezuela
Marriage Consensual union
1981 1990 2001 1981 1990 2001
Education level
Less than primary 5.71 5.10 3.54 6.72*** 5.75*** 4.29***
Primary 4.76 4.46 3.14 5.42*** 5.16*** 3.79***
Secondary 4.71 4.55 2.93 3.81 4.73 3.50*
University 3.42 4.27 1.83 0.86*** 3.80 0.00***
Labor force status
Employed 3.87 3.85 2.58 4.39 3.98 2.29
Unemployed 3.53 4.16 3.21 3.74 4.20 3.45
Inactive 5.40 4.88 3.57 6.60*** 5.81*** 4.30***
Area
Rural 6.71 5.41 3.60 7.26* 6.14*** 4.38***
Urban 4.87 4.47 3.25 5.87*** 5.19*** 3.80***
Home ownership
Owner 5.14 4.55 3.18 6.32*** 5.37*** 3.74***
Other 4.90 4.73 3.50 5.90*** 5.47*** 4.25***
*: p<0.05; **: p<0.01; ***: p<0.001. Each test compares the estimated TFR for consensual union to the corresponding estimate for
marriage.

36
FIGURE 1 Estimates of the age -specific fertility rates of women aged 15- 49 living in selected Latin
American countries according to their conjugal situation, 1980 -2010.
0.000.100.200.300.40Argentina 1980 Argentina 1991
0 .1 .2 .3 .4Argentina 2001
0.000.100.200.300.40Bolivia 2001 Bolivia 2001 Bolivia 2001
0.000.100.200.300.40Brazil 1980 Brazil 1991 Brazil 2000 Brazil 2010
0.000.100.200.300.40Chile 1982 Chile 1992 Chile 2002
0.000.100.200.300.40Colombia 1985 Colombia 1993 Colombia 2005
0.000.100.200.300.40Costa Rica 1984 Costa Rica 2000
0 .1 .2 .3 .4Costa Rica 2000
0.000.100.200.300.40Cuba 2002 Cuba 2002 Cuba 2002
Marriage Consensual union Not in union

Source: Census data from IPUMS. Estimates from the own -children me thod. Vertical lines at age 20, 30 and 40.

37
FIGURE 1 Estimates of the age -specific fertility rates of women aged 15- 49 living in selected Latin
American countries according to their conjugal situation, 1980 -2010. (Continued).
0.000.100.200.300.40Ecuador 1982 Ecuador 1990 Ecuador 2001 Ecuador 2010
0.000.100.200.300.40Mexico 2000 Mexico 2010 Mexico 2010
0.000.100.200.300.40Panama 1980 Panama 1990 Panama 2000 Panama 2010
0.000.100.200.300.40Peru 1993 Peru 2007 Peru 2007
0.000.100.200.300.40Uruguay 1985 Uruguay 1996 Uruguay 2006
0.000.100.200.300.40Venezuela 1981 Venezuela 1990
0 .1 .2 .3 .4Venezuela 2001
Marriage Consensual union Not in union

Source: Census data from IPUM S. Estimates from the own -children method. Vertical lines at age 20, 30 and 40.
Values over 0.40 for Uruguay are not depicted.

38
FIGURE 2 Point estimates and 95% confidence interval of the total fertility rates of women aged 15 -49
living in selected Latin A merican countries according to their conjugal situation, 1980 -2010.

Source: Census data from IPUMS. Estimates from the own -children method and Poisson regression.

39
FIGURE 3 Point estimates and 95% confidence interval of the total fertility rate of women aged 15 -49 living in selected Latin American
countries according to education level by conjugal situation, 1980 -2010.
MC
MC
MCM
CMC
MC
M
CMCMC
MC
MCM
C2.03.04.05.06.07.0
1980 1991 2001Argentina
MC
MC
MC
MC
2.04.06.08.0
2001Bolivia
MC
MCM
CM
CMC
MC
MC
MCMC
MC
MC
MCMC
MC
MC
MC 2.03.04.05.06.07.0
1980 1991 2000 2010Brazil
MC
MC
M
CM
CMC
MC
MC
MCMCMC
MC
M
C
1.02.03.04.05.06.0
1982 1992 2002Chile
MC
MC
MC
MCMC
MC
MC
MCMC
MC
MC
MC
1.02.03.04.05.0
1985 1993 2005Colombia
MC
MC
MCM
CMC
MC
M
CM
C
0.02.04.06.0
1984 2000Costa Rica

Source: Census data from IPUMS. Estimates from the own -children method and Poisson regression. From left to right, for each census, the education
levels are less than primary completed, primary completed, secondary completed, and university completed.

40
FIGURE 3 Point estimates and 95% confidence interval of the total fertility rate of women aged 15 -49 living in selected Latin American
countries according to education level by conjugal situation, 1980- 2010. (Continued).
MC
MC
MC
MC
1.01.52.02.53.0
2002Cuba
MC
MC
M
CM
CMC
MC
MC
MCMCMCMCMCMCMCMC
MC
0.02.04.06.08.0
1982 1990 2001 2010Ecuador
MC
MC
MC
M
CMC
MC
MC
MC
2.03.04.05.0
2000 2010Mexico
M
C
MC
MC
MCMC
MCM
C
MCMC
MC
MC
MC
MC
MC
MC
MC
0.02.04.06.08.0
1980 1990 2000 2010Panama
MC
MC
MC
MCMC
MC
MC
MC2.03.04.05.06.0
1993 2007Peru
MC
MC
M
CM
CMC
MCMCMCMC
MC
MC
M
C 0.02.04.06.08.0
1981 1990 2001Venezuela

Source: Census data from IPUMS. Estimates from the own -children method and Poisson regression. From left to right, for each census, the education levels are
less than prima ry completed, primary comp leted, secondary completed, and university completed. Uruguay is omitted from this figure because of inconsistent
estimates in 2006.