REGISTERING A MARRIAGE AND HAVING A CHILD IN RUSSIA: AN ANALYSIS OF INDIVIDUAL BIOGRAPHIES BASED ON VITAL STATISTICS DATA *

The paper looks at the closeness of the link between two events in the lives of Russian women: conception and marriage registration. The authors use the disaggregated 2010 data of current vital statistics on births for selected regions of Russia to analyze the relationship between marriage registration date and the date of conception, as well as age-related and regional features of this relationship. Special attention is given to the phenomenon of registering out-of-wedlock births based on the parents’ joint application as well as to the association between this type of conduct and the parents’ age. Using 2010 Census micro-data the authors examine the change of women’s family status during the first five years after the birth of the first child. The results of the study reveal that in Russia the widespread practice of registering marriages during the second trimester of pregnancy still persists. This is most typical among young people getting married for the first time. Together with the relatively high break-up rates of marriage among couples with children, this supports the interpretation of this kind of marriage as a marker of traditionalist demographic behavior.


INTRODUCTION
With the modernization of demographic behavior comes a weakening of the link between key events in the spheres of marriage and reproduction.In other words, the beginning of cohabitation and the registration of marriage can occur far apart in time.The official registration of marriage no longer leads rapidly to procreation, while pregnancy does not always prompt registering the marital union.
Data from Russian demographic surveys show increasing length of cohabitation before starting a registered union, growing number of couples not registering their marriage at all, and, accordingly, a significant proportion of out-of-wedlock births [Zvereva, Arkhangelsky 2011;Rosstat 2013].At the same time, a weakening of the link between the official registration of marriage and starting a family can lead to another correlation: between the time of marriage and the onset of pregnancy (often planned).In this case, the registration has a predominantly legal purpose and is regarded by the couple as a rational step.
Shotgun weddings, i.e. weddings arranged due to the occurrence of an unplanned pregnancy, have always been popular in Russia, with their heyday in the late Soviet period.The low age of first sex, coupled with the limited availability of efficient modern contraception and a complete lack of sexual education, caused frequent conceptions at young reproductive age.Social and, even more, institutional norms of the time made it necessary to look for a way out.The 1944 ordinance of the Presidium of the Supreme Soviet of the USSR prohibited the establishment of paternity of out-of-wedlock children who were deprived of the right to take the name of their biological fathers.According to experts, the introduction of this provision de facto reintroduced the concept of "illegitimacy".This had a significant impact on the status of women and children concerned, and led to a rise in the number of marriages registered after and as a result of pregnancy, so-called shotgun weddings.
In the years since the adoption of the post-perestroika family law, marital and family behaviors of Russians have changed significantly.Unregistered partnerships are by now a socially acceptable behavior; the beginning of partners' cohabitation is far from always tied to the official marriage registration; and marriage and family trajectories in general have become more diverse.A gradual development of contraceptive culture leads to a reduction in the number of abortions and, apparently, unplanned births.The shift in family policy towards the defense of children's rights, the fight against child poverty, and the overcoming of child abandonment have shaped the understanding that both parents can and should share the responsibility for the children born to them, regardless of whether they live in registered or unregistered unions.This is evidenced by the attempts to develop and implement such reforms as, for example, alimony laws.Thus, we can see a weakening of both social pressure and legal incentives for the registration of marriage, even when the partners have children.Have these changes affected such a widespread practice as shotgun weddings, and can we say that this phenomenon is disappearing, or has already disappeared, from the life of Russians?
The links between matrimonial and reproductive behaviors are actively discussed by Russian demographers, but little attention is paid to the topic of how conception affects the timing of official registration of marriage.Traditionally, research was focused on the prevalence of nonmarital births, and then the factors influencing its dynamics.Thus, G. Bondarskaya noted the rapid increase in the prevalence of extramarital births in the 1990s (from 14.6% in 1990 to 25.3% in 1997), interpreting it as modernization of the family and demographic behaviors related, among other factors, to weakening of traditional social norms [Bondarskaya, Darskiy 1990;Bondarskaya 1999].Conversely, M. Klupt showed a direct relationship between the prevalence of extramarital births in Russian regions and social disadvantages, as expressed in the low life expectancy and high levels of mortality among men of working age [Klupt 2010].
M. Tolts made the first attempt to use birth certificates to identify premarital conceptions [Tolts 1974].In this paper based on 1966 data from Perm, the author found that up to 35% of all births to women aged 20-24 resulted from premarital conceptions.In a later study, he demonstrated growth of this proportion during the ensuing 15 years [Tolts, Oberg 1983], which was associated with the simplification of procedures for marriage and divorce resulting from changes in family law.
In recent decades only few papers on this subject have been published, all focusing on estimating the proportion of non-marital conceptions.M. Tolts, O. Antonova and E. Andreev have shown that, in the early 2000s, extramarital conceptions in Russia accounted for about half of all births and about two-thirds of first births, and this proportion varied across regions [Tolts, Antonova, Andreev 2005;2006].Analyzing birth statistics, E. Ivanova and S. Zakharov conclude that by the beginning of the 21st century growth of the share of premarital conceptions ceased and decline of the age at first marriage generated by this factor stopped [Zakharov, Ivanova 2001].
Finally, the latest publication of this type is the article by E. Churilova and V. Chumarina [2014] which came out as our text was being finalized.The results presented in the paper prove that, during the period between the last two population censuses, among the youngest group of women (aged under 18) the proportion of non-marital conceptions has remained virtually unchanged, remaining at 90% or more.Our data for 2010 confirm the conclusion of other authors [Churilova, Chumarina 2014;Tolts, Antonova, Andreev 2005] that, in the general population, the proportion of births of all orders arising from pre-marital conceptions (which is not the same as non-marital births) decreased for all births, from 54 to 44% in 2002-2010, and for first births, from 68 to 60% [Churilova, Chumarina 2014-the author's data for 2010, and for 2002data from Tolts, Antonova, Andreev 2005].These authors interpret the continuing high prevalence of shotgun marriages as a conscious behavior of partners and choice of a strategy of "getting married only if pregnant" or rejecting contraception soon after deciding to get married.In this paper, we continue this discussion and give a number of arguments against this interpretation.Using an almost identical empirical base, namely, individual data from current registration of births, we dwell on several topics covered in the article by E. Churilova and E. Chumarina [2014].Our method of data analysis and visualization makes it possible to go beyond a descriptive approach to the evaluation of premarital conception.We analyze regional and age characteristics of marriage registration as linked to the conception of a child.Due to the nature of the data, we do not consider partnership relations before marriage and, consequently, cannot assume that the birth was the result of a conscious decision by the parents.Nevertheless, an analysis of the demographic portrait of parental couples and a detailed study of the age profiles of men and women who follow different strategies of family formation lead us to somewhat different conclusions than in this previous study.

STATISTICAL DATABASE FOR THE RESEARCH
The calculations presented in this paper are based on anonymous individual data of current birth statistics. 1 We had access to 2010 records from registry offices in ten regions of the Russian Federation.From 1998 to the end of 2013, the federal regulation did not impose recording the child's birth order, 2 although in some regions registration by birth order continued.Yet within this 1 Unless otherwise noted, all tables and figures represent the authors' calculations based on 2010 marriage and birth data from registries."Unmarried women" in tables and graphs refers to women who are not in a registered marital union. 2 To provide the data for the monitoring of the Russian President May Decrees (2012), the Decree of the Government of the Russian Federation N 1049 from 21 November 2013 approved the new Rules of reporting civil registration data by the offices of the Registry (ZAGS) to the Federal Service of State Statistics (Rosstat).These rules reintroduced the practice of collecting the data on the birth order of newborns.Moreover, the list of indicators was enlarged to include educational attainment and employment status of the mother.WWW.DEMREVIEW.HSE.RU research, we needed to separate first births from subsequent ones, which restricted the sample to regions with such data.As a result, our database was populated with data from the following seven regions of Russia: Primorsky Krai, Stavropol Krai, St. Petersburg and Leningrad Region, Moscow Region, Sverdlovsk Region and the Republic of Tatarstan.The dataset contains 301.3 thousand records of births made by official registries during 2010.These births occurred from late 2009 (0.7% of records) to the end of 2010 (99.3 % of records). 3n accordance with the Federal Law4 , a birth certificate includes the date of birth, the date of birth of the mother, the marital status of the mother, the date of marriage, the date of birth of the child's father, the ethnicities of the mother and father (optional), as well as some other information (single or multiple birth, live birth or stillbirth, administrative data).On the basis of these data, for each registered event we were able to calculate the age of the parents at the time of the child's birth, to determine the length of the marriage -if it had occurred -and to correlate the time of conception of the child with that of the official registration of the partnership.
Marriages and births certificates may be issued not only in the place of official residence, but in other localities as well.For example, in 2011 (similar data for 2010 are unavailable), an average of 97.7% of births in Russia were registered in the locality coinciding with the place of residence of the mother.This proportion is expectedly lower in large cities.For example, in St. Petersburg, this proportion comes down to 87.6%, with the remaining births being given by visitors arriving mostly from Leningrad region.Inasmuch as our work does not, with the exception of St.
Petersburg and its region, analyze the data by type of locality, such a bias should not have a material impact on the results.
Special attention is given to information on the mother's ethnicity.Incorporation of these data into the study of the distinctive features of marriage and family behavior in modern Russia could, in our view, provide interesting results.Due to the fact that ethnic affiliation of parents is indicated on the birth certificate only at the request of the applicant, in many regions this line is often left blank.In particular, in 2010 the ethnicity was almost completely missing in birth certificates issued in Primorsky Krai, Stavropolsky Krai and Sverdlovsk region; even in Moscow region this section was often left blank (30.7% of all certificates).In our analysis, therefore, we opted to use the data on the mother's ethnicity only for the Republic of Tatarstan.In the birth registration dataset for this region, 30.8% of mothers are recorded as Russian, 45.6% as Tatar and another 5% as belonging to other ethnic groups; in 18.6% of cases, ethnicity is not specified.We assume that, between Russians and Tatars, there are significant differences in preferences regarding the timing of marriage registration, so in most cases we considered them separately.
In addition, we relied on census data on first births preceding the 2010 Population Census in the regions selected above.This allows the use of two sets of data on the same population: from birth certificates and from census questionnaires Census data provided additional information on the marital status of women who were not officially married at the time of delivery.

FEATURES OF MARRIAGE AND FERTILITY IN SELECTED REGIONS
We start with a brief overview of the demographic portrait of the population under research.Despite the fact that, to date, we already have official aggregated statistics for 2011-2012 and sometimes even 2013, we analyze all indicators as of 2010, so as to be able to directly link them with the registry office data on which all our own calculations are based.The exception is information about the age at marriage, which was not processed by Rosstat for 2010 and therefore has been taken for the closest available year -2012.2010 marriage rates in the selected regions are close to the average level for Russia: about 11.5 marriages per 1,000 persons of marital age (16-69 years).The largest variation in this indicator is observed between St. Petersburg (13.3 per 1,000) and its province, Leningrad Region (8.3 per 1,000), which can be attributed to administrative reasons, that is, to the desire to register marriage in a big city.A similar difference is also apparent between Moscow and Moscow Region, though it is not considered in the present work.
The average age of brides and grooms has increased steadily since the early 1990s [Naselenie Rossii 2013].This applies to second or subsequent marriages and even more so to first marriages.In the Republic of Tatarstan and Stavropol Krai, the highest proportion of brides are aged 18-24, while St. Petersburg has the highest percentage of women who marry between the ages of 25 and 34.Other regions show a fairly homogeneous distribution of grooms and brides by age.Age-specific marriage rates suggest that the only region with a relatively high marriage age is St. Petersburg (Figure 1).The total fertility rate (TFR) at 1.116 children per woman is lowest in Leningrad region while current fertility is above the national average (1.567 children per woman in 2010) in Sverdlovsk region (1.668 children per woman) and the Republic of Tatarstan (1.601 children per woman).The TFR in other regions in the sample is low varying from 1.387 to 1.492.The oldest age profile of fertility (all births by order collapsed) among the observed regions is found in Stavropol region.Then come St. Petersburg and Moscow Region, while in the remaining regions the contributions of age groups 25-29 and 20-24 are almost identical (Figure 2).
The proportion of children born out of wedlock ranges from 20.4% in the Republic of Tatarstan to 30.7% in Primorsky Krai (Table 1).The specificity of Russia consists in that birth out of wedlock still remains a marker of adverse conditions and not of "modern" demographic behavior.In other words, the registration of births outside formal marriage in more than half of the cases suggests the absence of partnership.On the national level, 24.9% of all births occurred outside formal marriage in 2010; less than half (11.1% of all births) were registered by joint request of parents.At the same time, the highest proportion of births outside marriage is traditionally found in rural and economically backward regions (for example, in the Republic of Tuva or Magadan Region).
WWW.DEMREVIEW.HSE.RU   Thus, for all selected regions the macro indicators demonstrate the typical Russian model of population reproduction: low average number of children per woman, high proportion of firstparity mothers and concentration of childbearing between ages of 25 and 30.

DATA FROM CURRENT BIRTH RECORDS: A DESCRIPTIVE ANALYSIS
We now turn to a descriptive analysis of our data.The dataset contains 301.5 thousand birth records, of which 52.7% are on first births, 35.5% on second births and 11.8% on higher-order births.The proportion of first-borns varies from 46.1% among the group of women6 from the Republic of Tatarstan who did not indicate their ethnicity to 59.4% among women from St. Petersburg (Table 2).Altogether, the data array contains records on 158.9 thousand firstborns.
73.6% of firstborns are born within wedlock.Traditionally, the highest proportion of children born out of wedlock is in the youngest age group, and the lowest in the age group of 25-29-year-olds (Table 3).
Among Tatar mothers, the proportion of children born out of wedlock is the lowest.The highest proportion of first births out of wedlock is observed in Primorsky Krai, where it reaches one-third (Table 4).Generally, we found that the information about the child's father for children born out of wedlock is often omitted in the data, and therefore it is difficult to evaluate the true proportion of births to mothers with no partner.
The age distribution of mothers of firstborns in St. Petersburg is significantly different from that in other regions (Figure 3) with peak fertility in the age group of 25-29-year-olds.The proportion of first children born to older mothers in St. Petersburg is also higher than in other regions, while the contribution of the youngest age group is lower and amounts to just 4.1%.Among the Tatars in the Republic of Tatarstan, women aged 20-24 contributed most to the total number of first births, while younger women contributed almost as few first births as those in St. Petersburg.Stavropol Region leads in the proportion of first births to mothers in the younger age groups.There the contribution 20-25-year-olds to the number of firstborns exceeds that of 25-29-year-olds by a factor of 1.6 -more than in any other region.Moscow Region is unique in that the contributions of these two age groups are virtually equal (37-38%).

Figure 3. Distribution of first births by age of mother in selected regions, per cent
However, we must realize that we are dealing with period fertility indicators, which do not necessarily yield correct estimates of cohort fertility.Delaying first births and aging of subsequent childbearing began in Russia in the 1970s and are currently unfolding.The age profile of fertility is changing gradually and fairly slowly.Currently fertility is relatively low, including the firstorder birth rates as well as the rates at older ages and all parities.At the same time, the youngest cohorts will likely have higher fertility in their late reproductive years than we observe among 35-40-year-olds now.Similar processes may happen in the field of marriage and family behavior: the transformation of attitudes towards registered marriages and births out of wedlock may be slow.Period age-specific marriage indicators show that young women have already adopted the new model of behavior, while older women tend to follow the old behavioral stereotypes.
We'll begin our analysis of the interaction of such events as marriage, conception and childbirth in the life path of Russians by focusing on first births.

EVENT ANALYSIS OF MARRIAGE, CONCEPTION AND FIRST BIRTH
The nature of the data makes it inappropriate to use the calendar time-scale because all events are concentrated in too narrow a time interval of one year (2010).Therefore, we decided to use a relative scale.A distinctive feature of the data is that we do not consider childless marriages.The level of eventual childlessness in Russia is still low: in 2010, the proportion of childless women in the cohorts of women coming close to the end of reproductive age was about 6.5% [Biryukova, Tyndik 2014].The proportion of "childless" marriages is likely to be different from the proportion of childless women, yet it is not possible to evaluate the former indicator.Such unions emerge and then dissolve without falling into the birth statistics, and their absence in the dataset shifts the estimates of interrelation between marriage and childbirth upward, that is towards stronger association.
As a reference point for the relative time-scale we have chosen the approximated date of conception of the firstborn child.On the charts below, this moment can be seen as a plateau between zero marks, the duration of which corresponds to one calendar month (i.e. the period from 8 to 9 full months before the exact date of birth of the child).This makes it possible to smooth out the impacts of differences in the duration of pregnancy.The observational countdown begins one year prior to conception and ends 9 months after conception, that is, in the first month after birth.Note that, in the initial data set, there were some cases of marriage which took place more than one month after the baby was born.In theory, these observations should not be included in the registry statistics, since the time period allotted by law for registering birth is exactly one month.All these cases happened in the same region and were excluded from the analysis.
We have considered only one (dummy) indicator of marital status: the woman is either in a registered marriage or she is not.Figure 4 shows the transition of women who in 2010 registered the birth of their first child from one marital status to another.Each curve represents the proportion of women in the region who are not in a registered marriage at each point in time.Accordingly, the proportion of women who are in a registered marriage is equal to 100% minus the proportion of those who are not.Time before conception is reflected as negative values on the x-axis while the duration of pregnancy has positive values.
The proportion of married women who are "one year prior to conception of their first child" is low in all the regions in the sample, and ranges from 16.2% in Primorsky Krai to 27.1% in St.
Petersburg.By the time of conception of the first child, the proportion of unmarried women gradually begins to decline.The greatest rates of decline are observed among Tatars in the Republic of Tatarstan.In this group, the timings of official registration and of conception are most closely connected.Moreover, as can be seen in Figure 4, this timing is rather traditional of the type "registration then conception."St. Petersburg shows the most modern model: in registered unions, the birth of the first child is often postponed, as evidenced by the lowest proportion of unmarried women in the entire period before conception.At the same time, the incidence of shotgun weddings remains similar in all selected regions.By the third month of pregnancy, the registrations of unions begin to snowball, and their rates peak in the sixth month (among Tatar women at the seventh).On average, every sixth marriage "with children"7 is registered after the fifth month of the bride's pregnancy.Note that, according to Russian legislation, the registration of the marriage takes place one month after submitting a marriage application to the registry office.Thus, the majority of shotgun weddings included in current statistics are likely to have been planned after pregnancy.
Altogether, 45% of all marriages under consideration take place when the woman is pregnant.On average for the selected regions, 26.4% of women are unmarried at the time of giving birth.There are, however, significant regional and national differences: among ethnic Tatars in the Republic of Tatarstan only 17.3% fall into this category, while among Russian women in the same region the figure is 27.9%.At the same time, 33.9% of women from Primorsky Krai give birth to their first child out of wedlock.As noted above, our data break off at this point, although it is likely that, in the first months after birth, more unions are registered.
Thus, it seems possible to identify three models of shotgun marriages that co-exist in Russia, while the dominant model varies from region to region.The modern model is characterized by a high proportion of marriages "registered in advance" and a moderate proportion of weddings celebrated during pregnancy.It can be assumed that such weddings occur often after prolonged cohabitation in an unregistered union.In our sample, this is most apparent in St. Petersburg where the age profile of fertility also corresponds to the modern type, with the creation of families being postponed until at least the age of 25.
In the Republic of Tatarstan, the traditional model is common, whereby conception often coincides with the registration of the union or occurs within the first year of marriage.Still, very few women are unmarried at the time of birth of their first child.Finally, Primorsky Krai is characterized by the so-called Soviet model, which is the most typical for shotgun marriages.They occur most frequently in the second trimester of pregnancy, while at the same time a relatively large number of women are unmarried at the time of giving birth.

AGE CHARACTERISTICS OF MARRIAGE AND BIRTH
A woman's age has a significant impact on how close is the connection between childbirth and marriage.Among the youngest women, practically none is married a year before conception (Figure 5).Only in Stavropol Krai and the Republic of Tatarstan (among Tatar women) does the proportion of married women in the 15-19-year-old group rise to 10% by the moment of conception (see Appendix Figures A-1 -A-5, which show the graphs for each age group by regions).It is even lower in other regions.After the start of pregnancy and until the eighth month, the rates of marriage registration among young people are very high (about 7 percentage points per month), and only in the last two months do they slow.Among Tatar women, 36% of the youngest women are unmarried at the time of first birth, while for the other regions the figures range from 46.7% to 57.7%.On average, and across all the regions herein considered, one in two young mothers is not married at the time of her first child's birth.The main reason for this is that pregnancies in the youngest age group are mostly unplanned.

Figure 5. Proportion of unmarried women by time (in months) spent before and after conception of the first child, by age, per cent
Between the ages of 20 and 24, women get married most actively.Moreover, for many of them, conception occurs close to the official registration of marriage.During the year prior to the onset of pregnancy, the proportion of unmarried women falls by almost 20 percentage points (from 87 to 68%).After this follows an avalanche of marriages in the first seven months of pregnancy.Throughout this period all regions show roughly equal rates of reduction of the proportion of unmarried women (Figure 12 of the Appendix).Overall, during pregnancy the proportions of unmarried women fall, on average by 30-40 percentage points.At the time of first birth, one in four women aged 20-24 is not married.
The picture of marriage is different for women who become mothers between the ages of 25 and 29.Starting with this age group, already one third of women live in a registered partnership "a year before conception", and the frequency of shotgun weddings falls sharply.Among 25-29year-olds, we still can see a gradual decline in the proportion of unmarried women in the year before conception, while in the older age groups it becomes virtually invisible.During pregnancy, approximately one in four mothers aged 25-29 gets married, while one in five women is unmarried by the time her child is born.Table 5 summarizes the results of this analysis.In general, among many women the registration of marriage and first birth occur in rapid succession.In terms of the concept of life course, this means that marriage and childbearing still form the status passage.Among women who give birth to their first child before the age of 20, about half go through both events (marriage and birth) within one year.Among mothers aged 25-29, the proportion of such women increases to 63%, and in the older age group it falls back to 50%.In the life of women who become mothers after age 30, these two events are farther apart.Among them, only about one-third go through both events in the same year.If a woman gives birth to her first child after the age of 25, then in one out of three cases it is a result of postponement in an already officially registered marriage.
For the analysis of matrimonial behavior of women giving birth to their first child past the age of 30, there is an acute lack of data on the total number of respective marriages and partnerships.Many women are not in their first partnership at this age, and in it a relatively late, by Russian standards, birth of the first child may be a consequence of a change of partner.The observed pattern is contradictory.On the one hand, about 55% of mothers in this age group get pregnant in wedlock, which is evidence in favor of postponement of childbearing in stable unions.On the other hand, at the time of childbirth more of them are unmarried than among younger mothers.This could be either a birth in a stable unregistered union, or a pursuit of the strategy of "having a child for herself".

Figure 6. Distribution of parental couples by the difference in the age of spouses and intervals between conception of the firstborn and registration of marriage, per cent
Unfortunately, we cannot separate, with available data, first marriages from subsequent unions, which would have been of particular interest with respect to men.We use the only available way of establishing a typology of partnerships: from the perspective of the difference in age of the spouses (Figure 6).The proportion of marriages contracted after conception of the first child (the first-born of the woman) is lowest when partners are of the same age.However, selecting (from all partnerships) couples of the same age and grouping them by age intervals makes it apparent that the proportion of marriages registered after pregnancy gradually decreases with age (Table 6).Concurrently, there is a steady increase in the proportion of marriages concluded one year or less before conception.The turning point, when weddings after pregnancy become less popular than weddings taking place several years before the birth of the first child, occurs in the 25-29 age group.
Thus, among women under 25 who marry men of the same age, shotgun wedding strategy is extremely popular, even dominant.At older ages it is less common, perhaps due to the growth of autonomy and independence: the 23-25-year mark can also be linked to large numbers of women completing higher education, entering the job market, becoming independent of their parents and, consequently, being pressured less by the older generation.This is also to some extent reflected in the increased proportion of births occurring outside of wedlock in couples where the wife is over 30.In addition, it is explained by the registration of second or consequent marriages in this age group.
The lowest proportion of marriages in which pregnancy occurs one year after registration or later is observed among couples with the biggest difference in the age of spouses.At the same time, out of wedlock births occur in this group more often, as compared to others.It can be assumed that a significant proportion of such unions are repeat marriages, at least for the older spouse, and to this is due their less traditional behavior.The second wave of marriages in which the first child is born begins at the age of 40.For men this is seen more clearly: the proportion of unions formed more than two years before conception starts shrinking past age 39 (Figure 7).Among women of this age the proportion of such unions is stable, but this can be at least partly explained by the dropping out of women who have already given birth to a first child.Spouses remaining married for a long time and in principle intending to have a child are by the age of 35 usually no longer childless.Among men in older age groups we are seeing the growth of new marriages contracted 1-2 years ago, or even less than a year ago, which is obviously a marker of remarriage.
Shotgun weddings are peculiar primarily to first births.However, a second birth too may lead to the registration of a marriage, particularly when the child is the first-born in a new partnership (Table 7).In the youngest age group of women who have given birth to a second child, the proportions are higher both of those not married at the moment of birth and of those whose marriage was registered less than a year before conception (compared with those giving birth to a first child).At the same time, the proportion of those registering their union during pregnancy is significantly lower (14% versus 44% for the firstborn).At older ages, from half to two-thirds of women give birth to their second child in a stable marriage (concluded more than a year before conception).However, 11-13% register their marriage during pregnancy.After age 25, the proportion of second births outside marriage is lower than that of first births.At older ages the gap reaches 15-17 percentage points.Thus, despite the increase in the contribution of remarriage to fertility, marriages triggered by the conception of a second child are not at all common.

REGISTERING BIRTHS OUT OF WEDLOCK WITH JOINT PARENTAL CONSENT
Indirect evidence of the presence of an unregistered union comes from the information regarding the father in the birth certificates8 in cases when the woman is not officially married.Altogether, our data set for first births contains 7,555 such cases, all of which were documented in just three regions: Stavropol Krai, Moscow Region and the Republic of Tatarstan.On average, they make up 33% of the records in Moscow Region and 50% in Tatarstan.
The age profile of mothers who have provided information about the father of the child born out of wedlock does not support the hypothesis that this situation is typical mainly for older age groups (Figure 8).The age structure of the group of mothers providing such information resembles an inverted U-shaped curve: it is lowest among the youngest groups, and highest in the most fertile reproductive ages, i.e. between 25 and 29.The extent to which marriage and birth of the first child are connected is influenced by the age of both spouses, as well as by age difference between them.Figure 9 shows that the rate of registration of marriages after conception is higher among young men than among young women.In other words, if both parents are equally young, they are more likely to have a shotgun wedding.Also rushing to get married are young men aged 20-24 who exhibit two types of traditional behavior: either getting married and soon conceiving a child or legalizing the relationship while the bride is pregnant.The rates of registration of marriages among men of all other age groups are very similar.
From the records on the fathers of children born out of wedlock in their birth certificate (in 3 regions) we can analyze the age difference between the father and mother.The smaller the difference in age the higher the proportion of women in a registered marriage (Figure 10).The proportion of unmarried women is over 10% in those cases where the father is older than the mother by 10 years or more, and in those where the mother is older than the father by 6 years or more.It is worth noting that the birth certificates do not contain information about the order of birth to the father (i.e.whether it is his first, second, third or a higher parity child).

Figure 10. Proportion of unmarried mothers who provide information on the father to the registry, by age difference between parents, per cent
Thus, a shotgun wedding is most typical format of union formation for couples, in which both the man and the woman are under 20.It is also more common in regions with high fertility at younger ages.The greater the age difference between partners, the more often a child's conception does not lead to registration of the union.

DYNAMICS OF THE MARITAL STATUS OF WOMEN AFTER THE BIRTH OF THE FIRST CHILD
As noted, current records do not provide information on the duration of an unregistered partnership before conception.Macro data do not contain information about partnerships before their registration as official marriages.This lacuna can be filled only on the basis of sample sociological surveys, but even the largest sample sizes are not suitable for a detailed analysis of this phenomenon.Moreover, our experience working with the biographies of partners makes us critical of the reliability of the data in retrospective surveys on them.
Nevertheless, we still have the opportunity to look in more detail at the marital status of women who have recently given birth to their first child, and to evaluate the status's dynamics based on microdata from the 2010 Population Census.To compare them with the results outlined in the previous parts of this paper, we focus on census data in the same seven regions.
Since, prior to this, we worked with the current records from registries also for the year 2010, we actually can analyze the census data on some characteristics of the same women in the same families in the year of birth of the child.However, since the census was carried out in October, it contains the data on only 71.7% of first children born in 2010 and in our dataset of current records.
The proportion of women in a registered marriage according to the census is in agreement with the proportion of births within marriage according to birth records (Table 8).The proportion of women recorded by the census as not being officially married is usually expected to underestimate the real share of unregistered partnerships.This is because the legal and factual concepts were mixed in the census question prompts, in particular, the very first question of the section used the term marriage, not partnership.However, in Moscow Region the proportion of women living in unregistered partnerships is higher according to census data than that of unmarried women who have left a record of the child's father (41% vs. 33%).In Stavropol Krai and in the Republic of Tatarstan, they are very close (38% and 42% in the former, 49% and 50% in the latter, respectively).
The level of education of women is closely related to their age and birth cohort.To illustrate the educational differences in marital status, we considered women aged 25-29 (Table 9).Less than 15% of women with higher education are not married in the year of birth of their first child.Unregistered marriages remain the prerogative of women with low levels of education, whose proportion is small in younger cohorts.Regional differences are small, and in particular, St. Petersburg does not show any deviation from this pattern.In other words, such a marker of the modern model of demographic behavior as unregistered marriages "with children" currently is not typical for Russia.
We can trace the changing marital status of women with census data.Among women who gave birth to their first child in 2006-2010, the proportion of those in unregistered marriages remained virtually unchanged at 11% during that five-year period.This suggests that, if the union is not registered by the time the baby is born, most likely it will not be registered at all.Within a five-year period, the proportion of separated and divorced women increased 4-fold -from 3 to 12%.The largest growth (from 2% to 14%) was revealed in the youngest age groups.In the oldest age group, the increase was about 5 percentage points (from 9.6% to 14.9%).Unfortunately, we cannot separate unions registered before and after conception and compare their stories.Earlier we mentioned that in stable informal partnerships registering marriage after conception might have mostly legal significance.However, the specific feature of the Russian society is that the importance of this factor is lower than in Western countries.In particular, this applies to divorce and to the fact that officially terminating a marital union does not provide any essential guarantees to either the father or the mother of the child.On the one hand, Russian judicial practice is biased against fathers and a child's custody is virtually always conferred to the mother.On the other hand, the widespread practice of child support (alimony) is based on verbal agreement rather than on a court decision.In other words, the settlement of disputes relating to the child after the dissolution of the union is often done informally rather than legally.In conjunction with other results of the analysis, this testifies to the fact that, in Russia, the registration of marriage during pregnancy plays mostly a traditional (ceremonial) role.

CONCLUSION
Shotgun weddings remain a significant aspect of Russian reality.Partnership and childbearing are also still highly correlated in other developed countries [Baizan et al. 2004;Musick 2007].In Russia, this close relationship is typical for men and women of all ages, but most of all for young couples in which both partners are under 25.Record high rates of marriage are observed among couples where the groom and bride are under 20.If the first birth occurs after a woman turns 30 these events are further apart from each other in her life.
From one third to half of out-of-wedlock births are registered by mutual request of the parents.Most often, this occurs in the middle of the reproductive age, more rarely in the youngest and oldest reproductive-age groups, in which there is a higher proportion of registration of births by request of the mother only.The higher the education of the woman and the smaller the age difference between partners, the more likely they are to opt for a registered union.
Currently, slightly more than half of mothers under 30 go through both events (conception of the first child and marriage registration) within one year.But as the modernization of demographic behavior progresses, the relationship of the timing of marriage with that of conception weakens.The Russian Federation is extremely heterogeneous in its demography: some regions have just embarked on demographic modernization, while others have advanced fairly far.In the most advanced regions of our sample (St.-Petersburg) the birth of the first child is postponed in a registered union more often than in other regions, while at the same time the popularity of shotgun weddings remains at the average level.This is combined with the lowest birthrate among mothers under 20.Leaders in the prevalence of out-of-wedlock births are the regions of the Siberian and Far Eastern federal districts (in this case, Primorsky Krai).
How has the significance of premarital conceptions changed over the past decade?In modern Russia, an indicator such as the proportion of first-borns conceived before marriage and born within wedlock increased (Table 10).While in the late Soviet period their prevalence depressed the average age at marriage, now this impact vanished.In the past, cohabitation and sex outside wedlock were not socially accepted.Registration of marriage during pregnancy could be considered then as an inevitable and the only way to legitimize the partners' cohabitation.Under modern conditions, premarital partnership has moved into the category of social norms, and the cause-and-effect link between marriage and conception weakened.
In the traditional society, typical building of the first family consisted in the sequence of the following events: registration of marriage, living together, having intercourse and conceiving a child.In the late Soviet period an alternative model emerged whereby the couple first engaged in sexual activity not involving cohabitation but eventually leading to conception which was followed by marriage and start of marital cohabitation.Currently, this model often includes premarital cohabitation with one or, sequentially, several partners; the resulting conception leads to the registration of marriage.Fertility of the youngest women has declined so greatly that mothers under the age of 20 have become a virtually marginal group.At younger ages, conception still contributes to legitimizing extra-marital cohabitation.As fertility shifts to older ages, the number of shotgun weddings tends to decline, but this phenomenon is unlikely to disappear completely.Premarital conception is not the same as unplanned pregnancy.There are four possible events, two of which are associated with contraceptive behavior.The first is contraceptive failure (unwanted pregnancy), the second is loose application of contraception sometimes associated with uncertain reproductive intentions of a couple.At present, the awareness of men and women of the effectiveness of contraceptives is significantly higher than in the Soviet era, and the frequency of contraceptive failure has decreased tremendously.The use of traditional methods of contraception and a loose adherence to the rules of application of modern methods are often the result of uncertain reproductive intentions.Qualitative sociological studies show that this is a fairly common behavioral tactic [Tyndik 2015].Two other events imply planned childbearing by the couple or by either of the partners.It is not possible to evaluate the prevalence of these variants, but conceptions planned by the woman alone are clearly common [Ipatova, Tyndik 2015].Finally, couples registering their marriage after a planned pregnancy may be more family-oriented than those who live in a childless marriage or who do not register the union even after the child is born.
To summarize, it is too early to say that in Russia premarital conception is the result of a conscious reproductive strategy of partners [Churilova, Chumarina 2014].Clearly in opposition to this conclusion is the high prevalence in the younger age groups (where birth rates are now declining) of marriages registered after pregnancy.In addition, when family planning becomes widespread, one should expect a jump in marriage registrations just before conception, not after it (because this is a more rational behavior).In such a case, the difference from the traditional model would be that the couple cohabits for some time before registering marriage.
Table 10.Proportion of extra-marital births in selected regions, by age of mother, per cent At older ages, marriage registration prompted by conception has rational (legal) and ritual aspects.Now, social penalties for cohabiting without a marriage license and for terminating an official union are almost a thing of the past.However, there are no visible socio-economic advantages of registering a union: firstly, because family support measures are focused primarily on the mother and child; secondly, because when a registered marriage is dissolved, its official status does not help to resolve the disputed issues with respect to child custody.Under these conditions, the registration of marriage become more of a rite reflecting more stereotypical than rational elements of behavior.The strategy of "registration only in case of pregnancy" reflects a classic motive for a shotgun wedding.Thus, it is no exaggeration to say that the registration of marriage after the onset of pregnancy remains a bright marker of traditional demographic behavior; moreover, there are no signs yet of this changing.

Figure 1 .
Figure 1.Age-specific marriage rates by gender in selected regions, 2012 5 Source: Authors' calculations based on unpublished current statistics data of Rosstat.

Figure 4 .
Figure 4. Proportion of unmarried women by the length of time (months) spent before and after conception of the firstborn in selected regions, per cent

Figure 7 .
Figure 7. Distribution of marriages by their occurrence relative to conception of the first child, by age of mothers and fathers, per cent

Figure 8 .
Figure 8. Proportion of mothers of children born out of wedlock who provide information to the registry on child's father, by age, selected regions, per cent

Figure A- 3 .
Figure A-3.Proportion of unmarried women aged 25-29, by the length of time (months) spent before and after conception of the firstborn, in selected regions, per cent