The Pennsylvania State University

The Graduate School

College of the Liberal Arts

BELOW REPLACEMENT FERTILITY IN EASTERN EUROPE:

A CASE STUDY

A Thesis in

Sociology and Demography

by

Cristina Bradatan

© 2004 Cristina Bradatan

Submitted in Partial Fulfillment of the Requirements for the Degree of

Doctor of Philosophy

December 2004

The thesis of Cristina Bradatan has been reviewed and approved* by the following:

Nancy Landale Professor of Sociology and Demography Thesis Adviser Chair of Committee

Robert Schoen Hoffman Professor of Family, Sociology and Demography

Gordon De Jong Distinguished Professor of Sociology & Demography and Director, Graduate Program in Demography

Michael Bernhard Associate Professor of Political Science

Glenn Firebaugh Department Head and Professor of Sociology and Demography

*Signatures are on file in the Graduate School.

Abstract

Below replacement level fertility currently characterizes most of the European countries. Researchers have tried to explain this particular situation; some of them have even characterized the situation as a “demographic shock,” following changes in political and economic systems, while others have interpreted the changes in fertility as a response to the pressure of a declining economy. Although there are significant changes in the attitudes regarding family and number of children all over Eastern Europe, and these attitudes play an important role in the fertility decline, the “second demographic transition” in Eastern Europe is not a very popular hypothesis, maybe because not all the characteristics of the second demographic transition (increase in number of children outside of marriage, high rates of cohabitation) are present in these countries. The present research analyzes fertility changes in Eastern European before and after 1990, focusing more on one country (Romania), and paying attention to the interrelationships between fertility and other demographic phenomena. The dissertation includes three chapters: Eastern European fertility, The politics of fertility in Romania and Influences of out-migration on fertility. Each of these chapters includes a short introduction, several sub-chapters and conclusions.

Table of contents

List of tables and figures………………………………………………………………v Acknowledgements……………………………………………………………………..vii

Introduction ………………….……………………………………………………..1

1. Eastern European fertility..………………………………………………...... 3 1.0. Introduction………………...... 3 1.1. Demographic transition in Eastern Europe…………………………………...4 1.1.1. Theory of demographic transition………………………………….….4 1.1.2. Eastern Europe from the demographic transition perspective……….. 9 1.1.3. Demographic transition in Romania…………………………………19 1.2.Short history of fertility in Eastern Europe during communist era……….….24 1.2.2. Women and Family in Socialist Thinking………………………...... 24 1.2.3. Policy context………………………………………………………...30 1.3.Fertility evolutions in Eastern Europe after 1990…………………………....41 1.3.2. Data…………………………………………………………………..42 1.3.3. Theories………………………………………………………………47 1.4.Conclusions…………………………………………………………….…….52

2. The Politics of Fertility and Their Effects in Romania………………...... 55 2.0. Introduction………………...... 55 2.1.History matters ………………………………………………………………57 2.1.1. Short political history ………………………………..….……………57 2.1.2. Communism, abortion and eugenics………………………..…….…..60 2.2. Data and Interpretations…..……………………………………….………...67 2.2.1. Censuses………………………………….………………………...... 67 2.2.2. Fertility surveys: 1967, 1974 and 1978……………………………….76 2.2.3. Individual data………………………………………………………...80 2.2.4. Postponement of fertility……………………………………………...88 2.3.Conclusions………………………………………..…………………………91

3. Influences of out -migration on fertility……………………………...... 92 3.0. Introduction………………...... 92 3.1. Migration and fertility. Interrelationships………………………………………94 3.2. East –West Migration in Europe – levels and theories. ……………………….100 3.3. Data…………………………………………………………………………….107 3.3.1. Census data ………………………………..….…………………...... 109 3.3.2. Survey data ……………...………….…………………..………….……114 3.4. Conclusions…………………………………………………………………...124

4. Conclusions…………………………………………………………………………126

Bibliography………………………...………………………………………...... 132

List of tables and figures

Table 1.1. Households in 1787…………………………………………………………..12 Table 1.2. Percentage never married in selected eastern European countries…………...13 Table 1.3. Age at marriage for regions from Eastern Europe…………………………..14 Table 1.4. Mean household size in the provinces of Bosnia-Herzegovina by religious affiliation, 1870………………………………………………………………...17 Table 1.5. Population by ethnicity, new and old territories, in 1930……………………20 Table 1.6. GFR rates for districts where the majority has a certain ethnicity, 1932…….21 Table 1.7. Natality rates for old Kingdom and Romania, 1910-1915…………………...22 Table 1.8. Proportion of legitimate-illegitimate births, 1932……………………………22 Table 1.9. Percentage of legitimate births in urban and rural settings, 1932……………22 Table 1.10. Rate of abortion ……………………………………………...……………..44 Table 1.11. Number of abortions, 1995……………………………………………...…44 Table 1.12. GDP in Eastern European countries, 1989-1992 …………………………..47 Table 2.1. Ethnic composition of Romania in 1930 and 2002…………………………..58 Table 2.2. Number of children born per 1000 women, by age groups…………………..67 Table 2.3. CFR from simulation and from 1977, 1992 censuses………………………..68 Table 2.4. Percentage of childless women………………………………………………68 Table 2.5. CFR for college educated women, 1977, 1992, 2002………………………..69 Table 2.6. CFR for high school educated women, 1977, 1992, 2002…………………...70 Table 2.7. CFR for elementary school educated women, 1977, 1992, 2002……………70 Table 2.8. Women with a college degree for all age groups, 1977, 1992, 2002………..70 Table 2.10. Women with elementary school education, 1977, 1992, 2002……………..71 Table 2. 11. CFR for Germans, Hungarian, Romanian and Gypsy women in 1992, 2002…………………………………………………………………………….73 Table 2.12. Number of children born per 1000 women in 2002, and simulated with 1992 ethnic structure………….………………………………………………………...74 Table 2.13. Distribution of births/abortions for first to fourth pregnancy………………82 Table 2.14. Distribution of births/abortions for the first pregnancy…………………….83 Table 2.15. Distribution of births/abortions for the second pregnancy…………………83 Table 2.16. Distribution of births/abortions for the third pregnancy…………………..83 Table 2.17. Distribution of births/abortions for the fourth pregnancy…………………83 Table 2.18. Percentage of women wanting another child………………………………90 Table 3.1. GDP in Eastern European countries……………………………….………...92 Table 3.2. Number of ethnic Germans living in Eastern Europe who migrated to Germany after 1990………………………………………………….……102 Table 3.3. Stock of immigrants from former Yugoslavian republics in European countries………………………………………………………………………………..103 Table 3.4. Difference in size between 1992 and 2002 ………………………………..109 Table 3.5. Number of women that are that should be alive in 2002, in the absence of migration………………………………………………………………………………..111 Table 3.6. Number of children ‘lost’ because of migration, for every 1000 women, 1992- 2002……………………………………………………………………………………112 Table 3.7. Estimated number of women-migrants ……………………………………112

v

Table 3.8. TFR and number of women age 15-49 years………………………………113 Table 3.9.Demographic indicators in the poorest and richest counties………………..116 Table 3.10. Fertility of Romanians, Hungarians and Gypsies in Romania…………….118 Table 3.11. Regression coefficients……………………………………………………120 Table 3.12. Regression coefficients for analysis, fertility……………………………..122 Table 4.1. TFR for selected Eastern European countries……………………………....126 Table 4.2. Proportional change of TFR, 1986-2002…………………………………...127

Graphic 1.1. GDP per capita, Eastern Europe versus Western Europe …………………9 Graphic 1.2. Population in Eastern and Western Europe, 1820-1950………………….10 Graphic 1.3. Values of index of fertility control for selected countries………………..16 Graphic 1.4. TFR , Albania, 1960- 2003……………………………………………….32 Graphic 1.5. TFR in Hungary, 1950- 2003……………………………………………..34 Graphic 1.6. TFR in Poland, 1950-2003………………………………………………..35 Graphic 1.7. TFR in Czechoslovakia, 1950-1990………………………………………37 Graphic 1.8. TFR in Bulgaria, 1950-2003………………………………………………38 Graphic 1.9. TFR in Romania, 1956-2000……………………………………………...39 Graphic 1.10.Number of abortions per 1 live birth, Romania, 1967-1998……………...39 Graphic 1.11. Evolution of TFR between 1950-2000 in Eastern Europe, Western Europe and North America……………………………………………………………………….41 Graphic 1.12. TFR during 1960-1998, selected Eastern European countries…………..43 Graphic 1.13. Cohort fertility for females born during 1945-1965 in Eastern European countries………………………………………………………………………………….46 Graphic 1.14: TFR in Western, Southern, Eastern and Northern Europe……………....53 Graphic 2.1. Rate of natality, Romania, 1899-1966…………………………………….60 Graphic 2.2. Percentage of urban population, Romania, 1930-2000…………………...62 Graphic 2.3. TFR in Romania, 1956-2000……………………………………………...63 Graphic 2.4. Infant mortality, 1960-1972……………………………………………….65 Figure 2.5. States for model 1………………………………………………………...…82 Figure 2.6. States for model 2…………………………………………………………...82 Graphic 2.7..Rate of transition to a first pregnancy ending with a birth………………..85 Graphic 2.8. Rate of transition to a first pregnancy ending with an abortion…………..86 Graphic 2.9. Rate of transition to a second pregnancy ending with a birth…………….86 Graphic 2.10. Rate of transition to a second pregnancy, ending with an abortion……...87 Figure 3.1. East-West main directions of migration in Europe, after 1990……………104 Figure 3.2. Historical regions in Romania……………………………………………..115

vi Acknowledgments

During my graduate studies at Pennsylvania State University, I have had the chance to work and take classes with many wonderful people, and I want to thank all of them for their support. Above all, Nancy Landale has been of greatest help to me all the way through graduate school and especially in the process of writing up the dissertation. My gratitude to her goes much beyond what could be said in this note. Glenn Firebaugh was my advisor during my first year at Pennsylvania State University. Working with Dr. Firebaugh was for me a very exciting intellectual experience. I took several courses with Professor Robert Schoen, and I also worked with him as a research assistant. His work as a teacher and researcher is very important for my formation as a demographer. I was very lucky to have the chance of meeting and taking classes with Michael Bernhard, who has the passion and patience of studying Eastern Europe. Gordon de Jong was my professor for the migration course, which proved to be a very helpful course for my dissertation. I worked with Dumitru Sandu while I was an undergraduate student in Romania and without his help and advice I would not ever been able to go and study abroad. He has always helped me with data and suggestions while studying at Pennstate. I am very grateful for his help.

The following people helped me to get data for this dissertation and I am grateful to all of them. Leo Morris, Florina Serbanescu and Wyndy Amerson from CDC, Atlanta, provided me with the database from Reproductive Health Survey, Romania, 1999 that I used in Chapter 2. Florina Serbanescu also helped me when I had troubles with the data. Mihaela Antal and Petru Muresan from Center for Health Statistics, Ministry of Health, Romania, gave me access to the fertility and abortion reports published after 1967. Professor Vasile Ghetau, Center for Population Studies, Romanian Academy, helped me to get data from National Institute for Statistics, Romania. Margarete Kulik, Max Planck Institute for Demography, allowed me to use information from her paper on population policies in Poland.

During my graduate studies I received financial support from a number of institutions. Without their help, I would have not been able to finish my dissertation. Graduate School of Pennsylvania State University awarded me a Graduate School Fellowship for studies in Demography for 2000-2001.During 2001-2002 I received financial support as a research assistant from the Population Research Institute, Pennsylvania State University. During 2002-2004 I received financial support as a research assistant from the Department of Sociology, Pennsylvania State University. During 2001-2002 and 2003-2004 Open Society Foundation, New York, awarded me a Soros Supplementary Grant for students from Eastern Europe. Hewlett Foundation, through the Population Research Institute, awarded me a dissertation support grant, during the summer of 2002.

vii

Introduction

Below replacement level fertility currently characterizes most of the European countries. Several researchers tried to explain this particular situation - some of them characterized the situation as a “demographic shock” following changes in political and economic systems, while others interpreted the changes in fertility as a response to the pressure of a declining economy. The idea of “second demographic transition” in Eastern

Europe is not very popular, maybe because not all characteristics of the second demographic transition (increase in number of children outside of marriage, high rates of cohabitation) are present in these countries. The poverty that affected many of the people from this region seemed to be the most powerful factor in explaining the low level of fertility.

However, during these last years, many changes occurred within Eastern Europe: some of the countries succeeded in having fully functional open market economies, recorded a good economic increase and many of them have friendly policies regarding families and children. Despite these important developments, fertility continued to be very low in the region and there is little hope for a fertility increase, at least in a short period of time.

My dissertation analyzes fertility changes in Eastern European before and after

1990, focusing more on one country (Romania), and paying attention to the interrelationships between fertility and other demographic phenomena. The dissertation includes three chapters called: Eastern European fertility, The politics of fertility in

1

Romania and Influences of out-migration on fertility. Each of these chapters includes a short introduction, several subchapters and conclusions.

The first chapter, Eastern European fertility, describes the fertility evolutions and

the main explanations of these trends in Eastern Europe. The goal of the chapter is to put

the last decade of Eastern European fertility evolution in a larger picture, trying to

highlight the factors that could lead to the large decrease after 1990.

The second chapter, The Politics of Fertility in Romania, analyzes the influence of

the 1966 law against abortion on fertility before and after 1990. The fertility trend during

1966-1989 in Romania was heavily influenced by the 1966 antiabortion law, so the

higher than replacement fertility during the period of 1967-1989 was not a natural

phenomenon, but a direct result of the political restrictions on using contraceptives and

performing abortions. If this is the case, the low fertility immediately after 1990 is not a

“strange” phenomenon, but an expected trend, only interrupted during 1967-1989. In

arguing for these ideas, I use data from the 1966, 1977, 1992 and 2002 censuses as well

individual level data from the Reproductive Health Survey, Romania, 1999.

The third chapter, Influences of out-migration on fertility, analyzes the

relationship between international migration and fertility, after 1990, in Romania.

Romania, as well as other Eastern European countries, is among the very few world

regions with high out migration and very low fertility. Using various sources of data

(Census data and surveys), I address the following questions: 1) are migration and low

fertility interdependent? and 2) does migration distort fertility ?

2

Chapter 1. Eastern European fertility

1.0. Introduction

In this chapter, I situate the last decade of fertility evolution in Eastern Europe in

a larger historical and social picture, highlighting the main factors that could have led to

the decrease in fertility that occurred after 1990. One of the important tasks of this

chapter is to situate the fertility in Romania after 1990 in a larger context, both

geographically (as part of Eastern Europe) and historically (during the 19th and 20th centuries).

The chapter is structured into four subchapters, each discussing a particular period of time in the demographic history of Eastern Europe. Demographic transition in Eastern

Europe discusses late 19th and beginning of 20th century fertility evolutions, Short history

of fertility in Eastern Europe during communist era presents the situation during 1944-

1989, and Fertility evolutions in Eastern Europe after 1990 investigate the changes that occurred in fertility after 1990 and the main explanations given to these evolutions. The last subchapter – Conclusions – summarizes the main ideas of this chapter.

3

1.1. Demographic transition in Eastern Europe

1.1.1. Theory of demographic transition

Demographic transition is a well-known theory in demography, and it is still considered to be a theory that is “alive” (Hirschman, 1994). The idea of a changing demographic pattern as societies modernizes appeared sometime at the beginning of the

19th century. Among the first who gave a shape to this theory were Notenstein(1953) and

Davis(1963). They observed that fertility and mortality decreased in most of the industrialized world and noted the relationships between these demographic phenomena and some other components of social life, such as modernization and economic development. The classical form of demographic transition states that industrialization and urbanization created the new ideal of the small family. An increasing level of women’s education, a lower pressure for traditional behavior, and a growing cost of rearing children were among the main factors which permitted the emergence of this new ideal of family. Another version of demographic transition theory - Davis’ theory of change and response (Davis, 1963) - explains the change in fertility as an effect of population pressure that resulted from a decrease in mortality. The population can respond to this pressure through migration, by decreasing fertility or using both of these methods.

Demographic transition theory, with its various versions, induced initially a lot of excitement among researchers, and for a number of years there were few voices criticizing it. That was a time when few empirical data were available, so it was almost

4 impossible to falsify it. However, during the 1960s, more data became available, and contradictory results appeared. Leasure, for example, found out that the decline of fertility in Spain was not a simple result of increasing level of urbanization, education and changes in occupational structure (Coale:xx). Other researchers showed that countries at various economic levels shared a similar pattern of declining marital fertility:

“They [Iskandar and Knodel] found a somewhat puzzling pattern in which the timing of the decline in marital fertility in England and in Hungary was only slightly different, despite the very different levels of education, mortality and stage of industrialization in the two countries. Another puzzling parallel in marital fertility occurred in Norway and

Rumania”. (Coale, xx).

In 1963 the Princeton Population Office, together with other research units from

Western Europe, began to put together and analyze data regarding the fertility transition in Europe. The conclusions of this project showed that the decline in fertility cannot be related to economic development in a straightforward manner because the interaction level between people from different regions plays an important role: “Some areas of

Europe that were similar with respect to socioeconomic development did not experience a fertility decline at the same time, whereas other provinces that were less similar socio- economically experienced nearly identical drops in fertility. The data suggest that this riddle is solved by examining cultural factors rather than economic ones. Areas that share a similar culture (same language, common ethnic background, similar life style) were more likely to share a decline in fertility than areas that were culturally less similar”

(Watkins, 1991). Some other factors, such as the level of economic inequality, were found to be important in explaining the decline of fertility: “the idea of family planning

5 seemed to spread quickly only until it ran into barrier to its communication. Language is one such barrier and social and economic inequality in a region is another” (Leasure,

1962; Lesthaeghe, 1977)

The period prior to the Princeton Fertility Project was, probably, the highest point attained by the demographic transition theory. Then, the fall began: several researchers began to criticize the theory from various points of view. Some adjustments to or replacements of demographic transition theory appeared, and some of them were quite successful. Despite these critics, demographic transition theory is still a useful tool because it synthesized well the demographic evolutions of the last hundreds of years.

Hirschman (1994) argues that demographic transition theory needs to improve by getting more “flesh”, that is, it has to be discussed more in particular contexts, taking in account the society’s life in a certain period. Even if all regions did not evolve in a similar way, all modern societies decreased their fertility and mortality levels. Modernization may not have played the most important role, maybe other factors are more important, but only studies done in particular cases can show which other factors did.

6

1.1.2. Eastern Europe from the demographic transition perspective

There is a generally human, permanent need for grouping things together in order to understand them. It is a truism to say that this need manifests itself in the study of

Eastern Europe, too. The 1990’s political changes in the communist block ruled out the nicely packed idea of a world divided into East and West, communist and capitalist, centralized and free-market economies. There is no longer a clear-cut manner of grouping together the former communist countries: some of them are now rich and became part of the European Union, whereas others are still poor and hardly surviving the transition to a free-market system. So some other ways of grouping these countries were to be employed. This must be one of the reasons why former “Eastern Europe” was replaced by “Central Europe”1 and “Balkan region” or “Southeastern Europe”. The idea of Central

Europe, directly related to the former Habsburg Empire, is relatively old, but was resurrected in the 1970’s (among others, by Czeslaw Milosz (1964) and Milan Kundera

(Rider, 1994)) in an attempt to make people aware of the significant differences between

Eastern Europe, on the one hand, and USSR, on the other hand. The “Balkan” nations seem to share only the fate of having been, for some hundred years, vassals of the

Ottoman Empire, and renown as a “barbarous” region especially during the Balkan war at the beginning of the 20th century.

1 In most of the articles and studies, “Central Europe” or Mitteleurope has a historical rather than a geographical meaning: not necessarily countries that are in the central part of Europe belong to ‘Central Europe”. However, a recent article (Caldwell & Schindlmayr, 2003), uses ‘Central Europe’ for the central part of Europe, without any relationship to the history of the concept. Here I use ‘Central Europe’ with its historical, not geographical meaning. 7

Although it is not always obvious which nations belong to which region, and most of the Balkan countries refuse to be considered Balkan, it is generally accepted that

Albania, Yugoslavia, Bulgaria, Romania and Greece belong to the Balkan region. Some studies have also included Hungary in the Balkans (Todorova, 1997), although in the case of Hungary there are good reasons to consider it as part of Central Europe. The inclusion of Greece in the same group with some former communist countries makes the discussion about the Balkans particularly interesting: Greece was never part of the communist block, so, its “likeness” to the other countries from the region could only be a result of having been part of the Ottoman Empire. On the other hand, it would be problematic to affirm that fifty years of completely different historical courses did not affect the alleged resemblance between Greece and the former communist countries from Balkan

Peninsula.

In the following, I will use ‘Eastern Europe’ for countries that belonged during the Cold War to Warsaw Pact2, because this was the initial meaning of this term through much of the Cold War (De Nevers, 2003). Some statistics include the USSR, Yugoslavia and Albania. In such cases, I was not able to distinguish between USSR and Eastern

Europe, but I put a note if they are included. Most of the statistics do not include Eastern

Germany after reunification.

Generally speaking, Eastern Europe, and the Balkans in particular, are rather unstudied areas from the point of view of historical fertility changes. The focus of the

Princeton Fertility Project was on Western Europe, because fertility began to decline

2 The following states were part of Warsaw pact: Bulgaria, Czechoslovakia, East Germany, Hungary, Poland and Romania (De Nevers, 2003). 8 there. The statistics published by the Princeton Fertility Project on Eastern Europe are mainly results of computations using a model for stable population (Coale, 1986:170).

For Romania, for example, the Princeton Project used data for 1899, 1930 and

1956 and these data were adjusted for under registration using the West model stable population, with life expectancy at birth equal to 45 (Coale:173). Definitions of the regions are missing from the conclusions of Princeton Fertility Project, and this is very important because regions often changed their boundaries in this part of the world.

Dobrogea, for example, a South Eastern region in Romania, had a different configuration before and after 1918, as a result of incorporating Cadrilater, a former part of Bulgaria, inhabited mainly by Bulgarians and Turkish origin population.

Graphic 1.1. GDP per capita, Eastern Europe* versus Western Europe

(1990 Geary –Khamis Dollars)

6000 Western Europe 5000

4000

3000 Eastern Europe 2000

1000

0 1820 1870 1900 1913 1929 1950

Source: Monitoring the World Economy, 1820-1992

*Here Eastern Europe includes Russia

9

Fertility evolved very differently in Eastern than Western Europe (the fertility decline

began later, but the decline was shaper), and the relationships between fertility decline

and other population phenomena are not easily understandable3, so the studies done in

Western Europe cannot be simply used for the Eastern part of Europe.

Historically, Eastern Europe was poorer than the Western part of the continent,

and the difference in wealth between the two regions increased during the first part of the

20th century (Graphic 1.1).

During the19th century and first part of the 20th century, population increased at a

higher rate in Eastern than in Western Europe, mainly because of a higher rate of fertility

in Eastern European countries and, although at the beginning of the 1800s, there were

fewer people living in Eastern Europe than in Western Europe, by 1870 Eastern Europe

was more populated than Western Europe (Graphic 1.2).

Graphic 1.2. Population in Eastern and Western Europe, 1820-1950*

350000

300000 Eastern Europe

250000

200000 Western Europe

150000

100000

50000

0 1820 1870 1900 1913 1929 1950

Source: Monitoring the World Economy, 1820-1992

3 For example Şerbu (2000) showed the fertility decline in Romania was not a result of a decline in nuptiality, because nuptiality actually increased after 1900. 10

*Here Eastern Europe includes Russia

In comparison with Western Europe, Eastern Europe began to record a decrease in fertility relatively late, and several explanations were advanced for why this happened.

One of them is related to poverty which resulted from lack of modernization; as Graphic

1 shows, Western Europe was always richer than the Eastern part of the continent. A second explanation refers to the different types of families in Eastern and Western

Europe. Although demographers talk about the existence of a ‘European pattern of family’, they always refer to the Western European system of family. Eastern European countries seem to have, historically, a different type of family relationships, similar to the

Asian type (Goody, 1996).

Hajnal (1982) argues that the Western European system of household formation is a very particular case, completely different than the system in all other parts of the world.

This system emerged earlier than the seventeenth century and it is characterized by late marriage for both sexes (over 26 for men and over 23 for women) and separate households for young married couples (separate from the parents’ household) (Hajnal,

1982 :452). Children from poor and landless families were sent, at a certain age, to be servants in another household; after they saved enough money they married and established households separate from their parents.

Eastern European populations are described as having a joint household system.

Men and women married early (mean age for men around 26 and for women around 21), the young married couple started life in the household of an older couple and, after a while, a household with too many couples incorporated would split into two or more households. This system of family helped fertility to remain at a high level, while the

11

Western European system was an impediment having too many children. Hajnal uses data from the 1787 Hungarian census as proof of the existence of such a system (Table

1.1.). It was probably, he says, because of service that Northwest Europe could operate with a balance between birth and death rates established at a lower level than prevailed in other preindustrial societies.

Table 1.1. Households in 1787

Population Married men Persons per household (000s) per household Hyngary Proper 6,085 5.22 1.05

Transylvania 1,372 5.03 1.03 Croatia 617 8.33 1.7 61 "Free Royal 485 4.45 0.84 cities" Total 8,559 5.28 1.06 source: Hungary(1960) apud Hajnal (1982)

Although Hajnal’s ideas are still popular, other researchers presented different arguments. Sklar (1970) discusses some of the problems in establishing a “family pattern” for Eastern European countries. The geographic criterion is not a very good one for classifying and describing populations, especially in regions where the populations are very different with respect to history and culture. Eastern Europe, she says, does not seem to have been a coherent and homogeneous region from a demographic point of view. If we look only to the percentage of never married women in various countries of

Eastern Europe, it can be seen that heterogeneity rather than homogeneity characterizes the region. The percentage of never married females is, for example, much higher in

Poland than in other Eastern European populations (Table 1.2).

12

Table 1.2. Percentage never married in selected eastern European countries.

Percentage never married Country Year 20-29 40-49 Pre-war Poland 1900 36.3 7.8

Romania 1899 13.6 3 Bulgaria 1900 14 0.9 Bosnia 1910 11.3 1.6 Serbia 1900 9.4 0.7

source: Sklar (1970)

Age at marriage also varies between countries. As Table 1.3 shows, regions such as Estonia, Latvia, Lithuania and Poland had a high age at marriage for males as well as for females, and their pattern is more similar to Sweden and England than to Romania and Serbia, for example

These data led Sklar to believe that the Eastern European countries did not have a single type of family system. At least two different types of families characterized this region. Balkan countries were characterized by the existence of zadruga, an “extended household of two or more nuclear families related by blood or adoption owning land, livestock, and tools in common, and sharing a common livelihood and residence” (Sklar,

1974 :235) while the other countries, outside the Balkan peninsula, had a more ‘Western’ type of family system.

13

Table 1.3. Age at marriage for regions from Eastern Europe

Province or Age at marriage Later becoming Date Country Males Females Estonia, Latvia and Lithuania Esland 1897 29.7 26.3 Lifland 1897 30.2 26.6 Kurland 1897 28.8 25.6 Kovno 1897 29.7 25.4 Poland Vistula 1897 25.8 23 Vilna 1897 27.9 24.2 Grodno 1897 26.3 22.8 Volynia 1897 25.2 21.5 West Prussia 1900 27.3 25.2 Posen 1900 27 25.3 Galicia 1900 26.9 23.3 Czechoslovakia Bohemia 1900 27.9 25.4 Moravia 1900 27.8 25.4 Silesia 1900 27.7 25.2 Slovakia 1891 24.6 21.3 Romania 1899 24.5 20.3 Bulgaria 1900 24.2 20.8 Bosnia 1910 25.5 20.5 Serbia 1900 23 20.1 Sweden 1900 29.5 27.5 England and Wales 1901 27.3 26.2 Finland 1900 27.9 25.6 Portugal 1900 27.4 24.9 Italy 1901 27.5 23.6 source: Sklar (1974)

Religion is considered an important factor in shaping one or another type of household in Eastern European countries: Balkans countries are, mostly, Eastern

Orthodox and Muslims but Czech and Polish are Catholics and Lutherans. For the

Catholics, the church competes with kin authority because, by asking those who became priests and nuns to not marry, family and marriage are devalued. On the other hand, Islam and the Eastern Orthodox Church were favorable to extended kinship because they do not

14 have a separate or celibate clergy or a Church organization that can rival kinship (Sklar:

237).Czech, Baltic and Polish regions are also characterized, at the beginning of the century, by a high number of out migrants (about 130,000 between 1900 –1914), much higher than Balkan countries’ emigration (around 8000). Sklar argues that this out migration also has an important impact on the supply of mates for marriage. A woman who could not find a mate would go to work in the town, so she will avoid or postpone marriage due to employment.

Botev (1990) focuses intensively on Sklar’s arguments when he discusses characteristics of the family system and fertility evolution in Balkan countries during the

20th century. He advanced a third hypothesis for why fertility declined later in Eastern

Europe (especially the Balkans) than in Western Europe. The most important feature of the fertility transition in the Balkan countries, he says, is a much faster decline in fertility than in Western Europe, despite the relative low economic development of the Balkan countries. The rhythm of fertility decline is somewhere between the pace of Western

European countries and countries from Asia (see Graphic 1.3). On the other hand, during the demographic transition, the Balkans had a high rate of population growth, much higher than in the rest of Europe (with the exception of Russia) - between 12-15 per thousand, and sometimes over 20 per thousand (Botev, 1990: 121). Botev thinks that, after the liberation of Balkan countries from the Turks’ domination, a lot of land began to be available to the native people (land which was before owned by the Turks). This permitted them, for a while, to have many children without being very poor. However, land began to be scarce after a while, and in the absence of an out migration tradition, people limited their births drastically so fertility dropped dramatically in a short period of

15 time. Although this explanation is interesting and very original, it is sustained only with data for Bulgaria, and it cannot be applied to other countries that did not have similar conditions (Romania, for example, where Turks did not have land because it was only a vassal country).

Graphic 1.3. Values of index of fertility control for selected countries

source: Knodel, 1977 apud Botev (1990 :111)

On several points, Botev does not agree with Sklar’s ideas. Firstly, he is not convinced that zadruga was a frequently encountered type of family in Balkans. He argues that zadruga was proven to exist only in a very limited territory of former

Yugoslavia and the western part of Bulgaria (Botev:112) and, even there, nuclear families were not rare. Hammel (Botev, 1990) showed that between 40 and 82 per cent of the families in different samples of the west Balkan population (regions where zadruga was considered to be predominant) were nuclear. Botev emphasizes that even if the people did

16

not live in the same house, the proximity of the families made the familial relationships

very strong. These strong kinships relations allowed people to marry young and have

many children but did not require the existence of zadruga.

Sklar’ arguments regarding religion (Muslim and Eastern orthodox) as playing an important role in maintaining higher fertility levels in Balkans in comparison with other

European regions are also rejected by Botev. Using data on the average size of households in Bosnia –Herzegovina (Table 1.4) he shows that the average household size is higher for Catholics than for Muslims in all regions, so it cannot be argued that a higher percentage of Muslims in the Balkans explains the higher fertility. However, he thinks that religion played an important role in the sharp decrease of fertility in the

Balkans during the 20th century. Orthodox Church, Botev says, has a much more liberal

ethical conception than the Roman Catholic Church by placing ‘primary responsibilities

on husband and wife for making conscientious decisions’ about birth control (Fagley,

R.M., 1967, apud Botev, 1990).This made the decision regarding abortion and contraception easier for Eastern Orthodox believers than for those who belong to the

Roman Catholic Church.

Table 1.4. Mean household size in the provinces of Bosnia-Herzegovina by religious affiliation, 1870

Province Muslim Orthodox Catholic Average (Sanjak) Banjaluka 10.6 14.8 12.1 13 Bihac 13.1 17.7 15.9 15.5 Herzegovina 10.5 12.5 12.1 11.1 Novi Pazar 10.2 15.3 - 12.3 Sarajevo 10.7 14.9 12.2 12.3 Travnik 11.5 16.6 14.5 13.9 Zvornik 4.7 13.1 139 8.9 Total 9.3 14.8 13.2 12.2

17

source: Botev, 1990

In 1986, Jean Claude Chesnais (Chesnais, 1992) published a heavy book called

‘The Demographic Transition’, where he describes the demographic transition in sixty seven countries, from 1720-1084. He argues that, for European developed countries; there are seven different types of demographic transition. Romania, Bulgaria and former

USSR are included in the same group, characterized by very high level of pre-transitional growth, a rapid mortality decline, fairly high maximum levels of natural increase and a very short demographic transition of 70-years (Chesnas:251). Central Europe (including

Czechoslovakia, Austria, Hungary and Poland) ‘experienced a plateau of maximal growth which was: (1) later than in northern Europe; 2) centered around 1900; and 3) relatively brief (only 10 to 20 years)’ (Chesnais: 231). Chesnais’ work is impressive, because he provides an immense volume of data, but it is not always very clear what sources he uses for these data. For example, he includes statistics about fertility in Romania during the

Second World War (1940-1945), although during that time there were no demographic or economic statistics recorded in Romania, as the statistical system did not work properly.

18

1.1.3. Demographic transition in Romanian provinces

Data referring to the evolution of fertility in Romanian regions are mainly

provided by the various censuses taken during the nineteenth and twentieth century. In

Valachia and Moldavia censuses were taken in 1860, 1899, 1912 and 1938. For

Transylvania there are two censuses that are mostly used: 1910 and 1930. However, not

all these censuses recorded the number of children in the household, and there are good

reasons to believe the data are not very accurate (Chesnais, 1992).

Most of the studies argue that the fertility decline in Romania began at the

beginning of the 20th century, after the First World War, and fertility attained replacement level around 1950. The decline of fertility followed the decline of mortality that began around 1880. However, until around 1920, the decrease in the natality rate (or in the estimated TFR) is less than 10% from the initial rate, so, following the Princeton Fertility

Project, the fertility transition had not begun. Other authors (Ghetzau, 1997) contend that the decline in fertility began around 1885, much earlier than is commonly believed.

The studies done in this area by now do not show how much heterogeneity existed in the Romanian territories. Romania, as all other countries from Eastern Europe, changed boundaries many times during 1850-1950. Actually, until 1859 there was no

Romania, there were only separate regions where the majority of the population spoke

Romanian: Wallachia, Moldavia, Transylvania, Besserabia and Bukovina. The first two regions were vassals of the Ottoman Empire, Transylvania and Bukovina were Austro-

Hungarian province and Bessarabia was part of the Russian Empire. In 1859 Moldavia and Wallachia were unified and formed Romania (usually called “Old Kingdom”). In

19

1918 –1919 the other three regions joined the first two. A part of Bulgaria (Cadrilater) became also part of Romania. However, during the Second World War Bukovina and

Besserabia became parts of USSR; Besserabia remained part of USSR until 1990 when it became part of Republic of Moldova.

Therefore, when somebody talks about “fertility transition in Romania” what is meant is hardly understandable: is it about nowadays-Romanian territory; is it about

Romania at the beginning of the century? Some studies refer to nowadays Romania

(Ghetau, 1997), other authors use for each period whatever Romania meant at that time

(Chesnais, 1992). Both methods have advantages and disadvantages: for the first method the most important disadvantage is the lack of appropriate data. The second method can give very biased result, because the changes in territory added or took apart very heterogeneous populations. To give only one example, after 1918 four new territories were added to the “Old Kingdom” of Romania: Transylvania, Bukovina, Besserabia and

Cadrilater. The new territories added many minorities too, so by 1930, 35% of the entire population were minorities, mostly Hungarians, Germans and Jews. Table 1.5 presents the composition of the Romanian population for “new “ and “old” territories in 1930:

Table 1.5. Population by ethnicity, new and old territories, in 1930

Ukra Roma Hung Germ Russi Bulga Region inian Serbs Jewish Turks Other nians arians ans ans rians s Romania 71.9 7.9 4.1 3.2 2.3 0.3 2 4 0.9 1.7 Oltenia 97.5 0.2 0.2 * * * * 0.2 * 1.5 Muntenia 93.4 0.8 0.5 * 0.2 * * 2.1 * 2.5 Old

territori Moldova 89.8 0.9 0.3 * 0.3 * * 6.5 * 1.8

d Dobrogea 44.2 0.3 1.5 * 3.4 0.1 22.8 0.5 18.9 8.7 Mixe Basarabia 56.2 * 2.8 11 12.3 * 5.7 7.2 * 4.5 Bucovina 44.5 1.4 8.9 27.7 0.9 * * 10.8 * 5.7 New territori Transilvania 57.6 29.1 7.9 * 0.1 * * 2.4 * 2.6

20

source: Romanian Yearbook, 1935-1936

As can be seen, the new territories had, in some cases, more than half of the population with a non Romanian ethnicity. This is not a problem for studying changes in fertility, if different ethnicities have a similar fertility behavior. However, this is not the case for Romanian provinces; as Table 1.6 shows, the ethnic profile of a region shapes fertility behavior too. Statistics from 1932 (Georgescu, 1940) show there are major differences in fertility between counties when the ethnic composition is taken into account. The average GFR for the departments with a Romanian majority is 190.4 for old territories and 175.0 for new territories (children per 1000 women). The average GFR for the departments with a Hungarian majority is 142.8 and for departments with German majority is 112.7. Interestingly, among the departments with a Hungarian majority, the lowest GFR is not found where there is the highest percentage of Hungarians, but where the proportion of Hungarians is around 60-70%.

Table 1.6. GFR rates for districts where the majority has a certain ethnicity, 1932

Districts with >50% GFR

Romanians 190.4 Hungarians 142.8 Germans 112.7 Ukrainians 132.4 Bulgarians 216.5 Turks 243.4 source: Georgescu, 1940

The various territories included in Romania during 1918-1930 were also at different levels of mortality and fertility when they formed “Great Romania”:

Transylvania, for example, had lower rates of natality than Old Kingdom had after 1900.

21

Table 1.7 presents the natality rates for Transylvania and Old Kingdom during 1910 -

1915:

Table 1.7. Natality rates for old Kingdom and Romania, 1910-1915

Old Kingdom Transylvania 1910 39.5 35 1911 42.5 35.8 1912 43.7 37.1 1913 42.3 34.3 1914 42.1 33.5 1915 40.4 22.7 source: Ghetau, 1997

The percentage of illegitimate births varies strongly too between different parts of

Romania as well as between rural and urban areas. For the western counties (Table 1.8) only 83% of the births were legitimate, but for the southern counties 92% of the births were legitimate (Georgescu, 1940).

Table 1.8. Proportion of legitimate-illegitimate births, 1932

Births to Other Total births married cases women

Total 100% 89% 11% Sothern counties 100% 92% 8% (rural) Western counties 100% 83% 17% (rural) source: Gerogescu, 1940

The percentage of births to unmarried mothers varies between urban and rural areas too

(Table 1.9).

Table 1.9. Percentage of legitimate births in urban and rural settings, 1932

Births to Other Total births married cases women Total 100% 89% 11% Rural 100% 91%22 9% Urban 100% 82% 18%

source: Georgescu, 1940

There were remarkable differences in the number of children between people with different professions. A statistic from 1932 shows that after ten years of marriage, a couple that worked in agriculture would have an average of 5.39 children, but a professional couple would have 4.61 (Georgescu, 1940). Of course, the number of children after 10 years of marriage is not the best measure of fertility, but it can be used for comparisons.

In conclusion, the Eastern European demographic transition had different characteristics for Balkan countries and Central European countries. Central European countries followed closely the Western European pattern of demographic transition

(Chesnais: 231), however the Balkan countries began the transition later, and the decrease in fertility was sharper than in Western Europe. The demographic transition was also shorter in the Balkans than in Western Europe. As data for Romanian regions show, a deeper study of what happened in Eastern Europe has to take into account that various populations living in the same region have different demographic transition patterns (as the case of Romanians, Hungarians, and other nationalities show) and that changes in regional boundaries, which occurred several times during the 19th and the first part of the

20th centuries, strongly affect the viability of studies.

23

1.2. Short history of fertility in Eastern Europe during the communist era

1.2.1. Women and Family in socialist thinking

Students of former socialist countries are astonished by the changes that arose in the family after the fall of communism. High rates of mortality and infant mortality, very low fertility and an increasing number of divorces seem to be the rule for the Eastern

European region. Furthermore, the number of children born in Eastern European countries strongly decreased, making the replacement of generations improbable. At the same time, thousands of children abandoned by their parents spend their life on the streets of Sofia or Bucharest and have to be taken in by state institutions. This was certainly not the situation before 1990. The big economic depression that affected Eastern

Europe after 1989 is usually given as an explanation for these changes it is argued that the anomy within the society is affecting the family too and determines distortions within it. Is this the most important cause? Is it not more probable that families will be stronger in difficult times? Life, especially economic life, during the communist era was, from many points of view, more difficult than nowadays and still the family seemed to be strong. Why did Latin American countries not follow the same pattern when they passed through an economic depression? Although fertility is decreasing in Latin America, there was never such a demographic shock as can be observed in Eastern Europe nowadays.

There is at least one more factor that has to be taken into account when trying to explain actual changes in the family in former communist countries: namely, that the state was very much involved in shaping family characteristics before 1990. ‘New man’ was part of

24 a ‘new family’, there was an ideal socialist type of family that the states wanted to make real, and the pressure on the population, manifested through laws and regulations, to attain the ideal was intense. When such pressure ceases to exists, all the anomical movements that were kept under control before become manifest with strong intensity for a period of time, until a new equilibrium is attained. What we might now be seeing in

Eastern Europe as anomy can be seen as a wave of freedom following artificial pressure.

Let’s think what this “pressure” meant. The new Communist man had to be less vicious than the average person in a capitalist society. Adultery was considered a sin, prostitution was forbidden and a divorced person in a communist society had few chances to be promoted. Anybody who found out about an extra-marital love affair of her (his) husband (wife) could go to report the fact to Party, and the guilty partner would eventually be punished. The Communist Party tried to maintain the stability of the family by any means; it established the ethical norms within the family as well as outside of the family. Was it good or was it bad? Was it too much involvement citizens’ private life? I do not intend to answer to these questions, as my focus is not the ethical problems of totalitarian societies.

These characteristics of family policies were a result of the Marxist school of thought. Marx, Engels, Rosa Luxemburg, Lenin, Stalin, and others shaped ideas regarding the ideal family and family relationships in a communist society. The central problem of Marx’ and Engels’ philosophy was not, of course, family but the nature – materialist or idealist - of the social world (Jessops: 22): “The first central issue in Marx’s work concerns the nature of social world and is expressed in the philosophical dispute between idealism and materialism.” The answer that Marx gave to this problem is that the

25

social world is a result of the mode of production and all the components of a particular

society would be shaped by them. “Religion, family, state, law, morality, science, art, etc.

are only particular modes of production, and fall under its general law”. By “mode of

production” Marx presents the economic base and structure of the society as having two

components: forces of production and social relations developed in the process of

production (Jessop:24-25). Following Hegel, Marx would say that family is a part of the

state, because the same persons are both citizens and participants in a family. As a part of

the state, the family is not independent at all, but it will borrow the characteristics of the

state and this is why family relations will be shaped by the mode of production existing in

a particular society: “Family and society constitute themselves as the state. They are

driving forces…They are entities determined by a third party, not self –determined

entities” (Marx:17). The origin of Family, Private Property, and State is an important piece in understanding Marx’ and Engels’ thoughts on the family and the relationship with society. The idea that women are equals with men is one of the most important and innovative of this book. There was a period in history, Engels argues, when matriarchal society was the rule: children belonged to the mother’s family, mother was the ruler of the house and the man was only an accessory of the household (Engels: 735). However, men managed to change their status and to push women into a subordinate place. The bourgeois family is as a place for slavery as the Latin root of the word “family” shows.

Familia [family], he said, “signifies the totality of slaves belonging to one individual” and “the modern family contains in embryo not only slavery but serfdom also, since from the very beginning it is connected with agricultural services” (Engels: 737). The monogamous family with women being subordinated to men is only a creation of modern

26 times. Before these times, women were more powerful: “The women were in great power among the clans [gentes], as everywhere else. They did not hesitate, when occasion required, to knock off the horns, as it was technically called, from the head of the chief and send him back to the ranks of the warriors” (Engels: 735). The children were reared up in common, and because of this, it was not important who the father of the children was. Prostitution did not exist because the life in common did not ask for the existence of women selling sex.

Private property was the main factor in building a monogamous family. Men began to be preoccupied with not raising children of other men, so they imposed great punishments on adulterous women:” In order to guarantee the fidelity of the wife, that is, the paternity of the children, the woman was placed in the man’s absolute power; if he kills her, he is but exercising his right” (Engels: 737). The patriarchal polygamous family is the transition from matriarchal family to monogamy, and monogamy is carrying all the

“sins” of the patriarchal family: “Monogamy was a great historical advance, but at the same time it inaugurated, along with slavery and private wealth, that epoch, lasting until today, in which every advance is likewise a relative regression, in which the well-being and development of the one group are attained by the misery and repression of the other”

(Engels: 739). In the bourgeois family men’s and women’s class positions are important; even in Protestant societies, where it was assumed that children would choose their bride by themselves, they have to choose within their own class. The marriage is a conventional one; a contract between two persons, an institution where the woman is heavily dependent on the man because he is the breadwinner. Both men and women try to regain their sexual freedom, so there are two are results of this hypocrisy of bourgeois

27 marriage: hetaerism and adultery: “Bourgeois marriage is in reality a system of wives in common and thus, what Communists might possibly be reproached with is that they desire to introduce, in substitution for a hypocritically concealed, an openly legalized community of women.” (Marx: 488). How did Engels and Marx try to solve this situation? On this particular point, they seem to have quite different opinions. Marx saw the communist society as similar to the matriarchal one: the children will be raised in common and there will be a community of women. Engels seems to be more prudent in answering to this question. He thinks that marriage has to be based on love, and when love disappears the parts have to have the right to separate: “If only marriages that are based on love are moral, then, also, only those are moral in which love continues… a definite cessation of affection, or its displacement by a new passionate love, makes separation a blessing for both parties as well as for society” (Engels, 1884: 751). He is very optimistic that in a society where marriages are based on love, lack of private property and free divorce, the institution of family will grow stronger: “What will most definitely disappear from monogamy, however, is all the characteristics stamped on it in consequence of its having arisen out of property relationships. These are, first, the dominance of the man, and secondly, the indissolubility of marriage. The predominance of the man in marriage is simply a consequence of his economic predominance and will vanish with it automatically. The indissolubility of marriage is partly the result of the economic conditions under which monogamy arose…” (Engels,1884: 750). He argues that a society where the equality between man and woman is guaranteed and the dignity of person as individual is preserved will result in a stable monogamous family and the two weaknesses of bourgeois family – prostitution and adultery – will be gone (Engels:

28

743): “The two eternal adjuncts of monogamy – hetaerism and adultery – therefore, play an almost negligible role here; the woman has regained, in fact, the right of separation, and when the man and woman cannot get along they prefer to part. “

In conclusion, both Marx and Engels focused more on equality between men and women within the family rather than on the stability of the family. In their writings, men and women as individuals rather than as a family are important, marriage has to be a mutual agreement and the freedom of husband and wife is what has to be preserved: if one of the parts does not maintain the agreement anymore, divorce is the best solution.

Lily Braun, a German socialist who wrote extensively about women’s status in socialism, did not agree with many of Engels’ visions on marriage. She talks about

‘marriage reformers [who] have cooked up their own marvelous theory [about]… union of two equal, intellectually and morally mature people… [as being] the happiest, nicest, and most enduring’ (Braun, 1987: 126). These arguments are based on the false assumption that intellectual understanding is identical to the sensual instincts. The economic independence of women, she says, will destroy one of the main pillars of marriage, because marriage is basically an ‘economic union for the purpose of producing legitimate heirs of the father’s property’ (Braun: 127). Braun was very aware of working mothers’ problems, and was one of the first to propose maternity incentives as full paid maternity leave and infant day care centers (Meyer: xiv).

29

1.2.2. Policy context

Fertility is both an individual and a societal problem (Demeny, 1986), societal, because the age distribution of a nation influences the country’s level of development, and individual, because, in the end, giving birth and rearing children is an individual/family problem. Following Berelson (1974), there are two types of fertility policies: the positive ones, which use only incentives in order to attain a society’s desired level of fertility, and the negative ones, which use all types of constraints against people in order to attain the desired goal. Positive policies in below replacement fertility countries offer incentives to families in order to increase or decrease fertility: money at births, monthly allowances for families with one or more children, maternity leave, benefits for mothers with many children and single mothers, paternity leaves, etc.

Negative policies are those through which governments limit access to contraception or abortion in order to increase fertility.

State socialist governments in Eastern Europe faced a difficult situation in addressing the problem of low fertility: on one hand, the low fertility rate called for pronatalist measures, on the other hand, Marxist ideology emphasized the importance of raising women’s status in the family and society. These two policy directions have different results. As many studies have shown, increasing the status of women leads to decreasing levels of fertility. It is thus complicated to discern the effect of one type of policy (pronatalist), while policies with an opposite effect (e.g. raising women status) are in place. Sometimes, the declared pronatalist policies could have an opposite effect. This

30

seems to be the case with policies that provide child care programs: with the availability

of child care centers, ‘women who want to work will have the opportunity to enter or

reenter the labor force much sooner; and the rewards of employment may compete

effectively with the satisfactions of additional children’ (US Commission, 1972a: 88,

apud Demeny, 1986)’. All of the former state socialist countries adopted positive fertility

policies and, as I will show in the following, they succeeded remarkably in raising

women’s status in society. Some of them tried to limit the number of abortions and access

to contraceptives in an attempt to raise the level of fertility. In the following section I

discuss briefly the policy context for each country. In the conclusions I highlight the

general characteristics of fertility policies in Eastern Europe.

Albania is often cited as a particular case within the Eastern European fertility

context. With an average of six children per woman at the end of World War II

(Falkingham and Gjonca: 310) and around three children per woman by 1990, Albania is

obviously an outlier with respect to fertility trends in Eastern Europe. However, fertility

is not the only field where Albania was ‘different’. Historically, one of the very poor

European countries4, Albania has three main religious groups: Muslim, Catholic and

Orthodox, with the great majority of the population (73% in 1945) being Muslim.

Situated in the Southern part of the continent, it was long known as the most isolationist country in Europe (Falkingham & Gjonca: 309). With almost all statistical data considered as a ‘state secret’, little demographic information on Albania was available before 1990.

4 GNI PPP in Albania - 3,810 US $ in 2001, average GNI PPP in Eastern Europe – 7170 US $ in 2001 (World Population Data Sheet, PRB, 2003) 31

Graphic 1.4. TFR , Albania, 1960- 2003

8 7 6 5 4 3 2 1 0 0 0 0 1 2 3 4 5 6 7 8 9 0 6 3 6 7 8 8 8 8 8 8 8 8 8 8 9 9 0 9 9 9 9 9 9 9 9 9 9 9 9 9 9 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2

source: Recent demographic developments in Europe, 1999 and World Population Data Sheet, 2003

From 1950 to 1990, fertility in Albania decreased from over six to three children per woman, and continued to decrease in the years after to a replacement level of 2 in

2003 (Graphic 1.4). During 1945 to 1990, under one of the most ‘isolationist orthodox

Communist regime’, fertility decreased to half of what was at the beginning, despite the pro-natalist policies implemented by the government. As in most of the former

Communist states, Albania had both positive and negative pronatalist policies. One set of positive policies was oriented toward decreasing the costs of bearing children by subsidizing children’s clothing prices, waiving costs of medical prescriptions for children

0-12 months and providing paid pregnancy leaves up to six months. Another set of positive policies addressed the dual roles of women as mothers and workers, by waiving fees for children age 3 to 6 placed in kindergartens and by building canteens at the

32 workplaces (Falkingham & Gjonca:315). Until 1991, Albania had one of the most restrictive pronatalist laws, with modern contraception devices not being either produced or imported and abortion being restricted to 30 well defined medical situations

(Falkingham & Gjonca:315, Dyme&Pine:61). However, abortion seems to remain an important method for family planning in Albania, with the abortion ratio5 being around

230 during 1980-1990 (Dyme&Pine: 55).

After 1989, the number of medical situations for which abortion was permitted was expanded, and by 1991 over 200 health situations were accepted (Dyme&Pine:61).

In June 1991, a directive of the Albanian Ministry of Health made abortion legal upon request up to 12 weeks, for social reasons up to 15 weeks and for special medical situations, without time limit (Dyme&Pine:62). Contraception was legalized and regulations regarding abortion were changed (Dyme&Pine: 56). In May 1992, family planning was adopted as a ‘fundamental right of all individuals’ (Dyme&Pine: 64). In

November 1995 the abortion law was adopted by the Parliament (Dyme & Pine: 62).

The status of women within Albanian society was traditionally very low.

Historically, women were controlled totally by their husbands, virtually imprisoned in the house (Dyme&Pine:56), without any property rights (Falkingham & Gjonca:315) and, for most of them, without any education. After World War II, when the Communists took over the country, 95% of women were illiterate and by 1950, the life expectancy of women was still lower than the life expectancy of men (Falkingham & Gjonca: 315).

Through various policies regarding education and the work force, by 1990, the percentage of women with higher education attained 51.3%, and the percentage of

5Abortion ratio is defined here as the number of abortions per 1000 women at childbearing age. During 1974- 1995, the number of abortions increased from 11422 (1974) to 32 588 (1995) (Recent demographic developments in Europe, 1999) 33 women in the workforce 45.1% (Falkingham & Gjonca:317). Although various policies contributed to a rise in women’s status, other aspects of life made women’s life harder in

Albania. Although the educational level and work force participation of women increased dramatically during the last 50 years, private life ‘remained patriarchal to a large degree’

(Dhmitri, 1985). In this poor society, a family with more than two children could not afford to live with only one salary, so women were required not only to work, but to take care of children, do house work and care for in laws (Dyme & Pine: 58), a triple burden not easy to deal with. It is not surprising that fertility decreased dramatically during the last ten years.

Hungary, a central European country, was one of the first to experience a natural decrease in population during peaceful times (in 1981, when there were 144757 deaths and 142890 births). Pronatalist measures have a long history in Hungary, beginning in

1924 when a three month (12 weeks) maternity leave was introduced the first time. After

World War II, in 1953, several measures were adopted which restricted access to abortion and contraceptives (David: 147). This was followed by an increase in fertility. However, after 1957, fertility began to decrease, attaining some of lowest levels during 1962-1965.

Graphic 1.5. TFR in Hungary, 1950- 2003

34

3 2.5 2 1.5 1 0.5 0 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2003

source: David, 1999, PRB, 2003

Several positive and negative pronatalist measures were initiated by the

Hungarian government after 1960: in 1965 family allowances were raised, in 1967 paid maternity leave was introduced and in 1970 stricter abortion regulations were passed

(Tomka, 2002:11). The expensive pronatalist population policies introduced after 1967 were successful and ‘more educated and higher professional categories showed an appreciable relative increase in fertility , as the policy intended’ (Demeny:350). Some researchers suggested that the pronatalist measures had an ethnic component too, with the government trying to stop the growth of some minorities, especially Roma (Kulcsar:11).

The effects of these measures were visible, with the TFR increasing after 1965 (Graphic

1.5.) although among the Eastern European countries, Hungary continued to have one of the lowest levels of fertility. As any other communist government, the Hungarian one tried hard to improve the status of women mainly by facilitating their participation in the labor market. By 1989, 80% of women age 15-45 (excluding students) were employed, and among women under age 40, 90% were in the workforce, although the inequalities between men and women continued to be present, women’s income being 25% less than that of men (Davis:162).

35

Poland has the largest population in Eastern Europe (38.6 millions in 2003) and a long history of high fertility (Graphic 1.6).

Graphic 1.6. TFR in Poland, 1950-2003

4 3.5 3 2.5 2 1.5

1 0.5 0 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2003

source: David, 1999, PRB, 2003

A country with a troubled history, Poland appeared as a state in Europe in 1918

after the unification of regions previously belonging to Russia, Germany or the Austro-

Hungarian Empire, and practically disappeared again during the World War II, being

firstly divided between Germany and Russia (1939) and then completely under the

German occupation (1941). Religion played and plays an important role in Polish

demographic history, for example, secular marriage was adopted only in 19456 and

divorce was made possible in 1946 (Fratczak, 2003:5). As incentives for having children,

maternity leave and allowance were introduced in 1924. During 1945-1974 several laws

changed the level and conditions under which a family with children was entitled to

receive birth allowances. In 1974 a birth allowance, in the amount of three times the

monthly allowance for a child was introduced and this law was again amended in 1990

6 Marriage Law – decree of September 25, 1945 36

(Fratczak& all, 2003:15-19). In the 1980s, a mother could ask for a three year unpaid maternity leave (Titkow:183). In 1956 abortion was made legal for women who were raped, who had medical conditions and or had difficult living conditions7 . This law was amended in 1993, social conditions not being anymore an accepted reason in asking for an abortion (Fratczak & all, 2003:5). In 1996-1997 the law was changed again, making easier for women with low income to get a legal abortion.

Czechoslovakia, divided in 1993 into Czech and Slovak Republics, recorded one of the highest decreases in fertility after 1985 (Graphic 1.7). Legal abortion, introduced in

1957, reduced the incidence of the third child, although the fertility rates remain high for the first and second child ( Sobotka & all:4).

Graphic 1.7. TFR in Czechoslovakia, 1950-1990

3.5 3 2.5 2 1.5 1 0.5 0 1950 1955 1960 1965 1970 1975 1980 1985 1990

source: Wynnyczuk&all, 1999

After 1945 a Family Allowances Act was introduced and state support was dedicated to families with children, with no restriction regarding income

7 Law of May 8, 1956 regulating the conditions of abortion 37

(Wynnyczuk&all:101). In the 1960s, worried that fertility would remain low for the future, the government raised the family allowance, especially for the second child, and provided a one year maternity leave. By the end of the1970s, 4% of the Czech budget was spent for family cash benefits and 7% on services for families with children

(Wynnyczuk&all:102). After 1945, while the Communist party had the power, the percentage of women in the labor force increased from about 37% (1948) to 90%

(1989), although men continued to have a 25% higher income than women

(Wynnyczuk&all:103)

Bulgaria recorded the largest decline in fertility during 1956-1966, as a result of the policies of rapid urbanization and forced agricultural collectivization (Vassiliev:71)

(Graphic 1.8).

Graphic 1.8. TFR in Bulgaria, 1950-2003

2.5

2

1.5

1

0.5

0 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2003

source: Vassiliev, 1999, PRB, 2003

Abortion was permitted in Bulgaria beginning in 1936, but only for serious medical reasons. During 1956-1968, abortion was made available on request for any

38 women who wanted it, but in 1968 some restrictions were introduced for married women who were childless or had only one child8. Family allowances for the third child and the birth allowance increased also in 1968, a measure aiming ‘to stimulate Bulgarian families to have at least three children and to discourage (primarily Turkish and Gypsy) families from having more than three children (Vassiliev: 76).In 1989 abortion became again available upon request.

In Romania, abortion was first legalized in 1956, the decision being based on the

Marxist idea that women are equal to men and women have the right to do what ever they want to their bodies. Between 1956 and 1966 fertility decreased strongly (see Graphic 8) attaining a below replacement level of TFR as early as 1963. In the meantime, the abortion rate recorded very high values, with the highest in 1966 (estimated to be around

5 abortions per 1 live birth). Interesting enough, during all these strong changes in fertility, the rate of marriage remained high and the age at marriage low, contradicting the usual pattern of declining fertility.

Graphic 1.9. TFR in Romania, 1956-2000

4.0

3.5

3.0

2.5

2.0 TFR

1.5

1.0

0.5

0.0

1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

8 In 1973 abortion was restricted to unmarried women. 39

Source: Demographic Yearbook, Romania, 2001

In 1966, worried by the decreasing fertility, the government emitted a law prohibiting abortion for a large female population9. The law was followed by a strong increase in fertility (see Graphic 1.9.) during the next three years, and it remained above replacement level until 1990. In 1990, the law against abortion was repealed and the immediate result was a decrease in fertility and a strong increase in the abortion rate (see

Graphic 1.10.). Fertility continued to decrease between 1990 and 2000, despite the various polices toward encouraging births.

Graphic 1.10.Number of abortions per 1 live birth, Romania, 1967-1998

3.5

3

2.5

2

1.5

1

0.5

0 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998

Source: Romanian Ministry of Health

9 There were several cases excepted: women with severe medical problems, older than 44 years or who had 4 or more (later on, 5) children alive. 40

1.3. Fertility evolutions after 1990

It is generally accepted that all modern populations followed the demographic transition pattern from high to low rates of fertility and mortality. When and at which limit the decrease in mortality will stop is one of the many unknown points about the demographic transition. Regarding fertility, many researchers hypothesized that replacement level is the lowest point that would be attained, and that fertility will remain more or less around that value. Unfortunately, this was not the case and nowadays most of the European countries have persistent below replacement fertility. Although the TFR values in Eastern Europe are not much smaller than the TFR values for Western Europe, what distinguishes them within Europe is a continuous decrease after 1990 (Graphic 1.

11). Several researchers tried to explain this particular situation of decreasing fertility in

Eastern Europe at the same time as constant or even increasing rates in Western European countries.

Graphic 1.11. Evolution of TFR between 1950-2000 in Eastern Europe, Western Europe and North America

4

3.5

3 Eastern Europe 2.5 Western Europe Northern America 2

1.5

1 1950-55 1955-60 1960-65 1965-70 1970-75 1975-80 1980-85 1985-90 1990-95 1995-00

Source: UN

41

Three main explanations were advanced to explain these fertility changes in

Eastern Europe: 1) some researchers characterized the situation as a “shock” (Eberstandt,

1994) or an anomical situation (Ranjan, 1999) following the changes in the political and economic system, 2), others argued for an economic explanation, emphasizing the increasing poverty that characterized the region after 1990 as a major factor in decreasing fertility (Carlson & Omori, 1998, Macura (2000), Cornia and Paniccià (1996)) and, 3) a smaller, but increasing group, argue for a cultural explanation of this ‘sudden’ (Philipov,

2001:10) decrease in fertility (Sobotka&all, 2001), a second demographic transition which would affect Eastern Europe after 1990. None of these perspectives are mutually exclusive, and, in fact, as Philipov (2001) observed, lack of adequate data makes difficult to decide which one can offer the best explanation. In the following I will present data that lead the researchers to identify fertility in Eastern Europe as a problem (section

1.3.1.), I will briefly describe the arguments used by the supporters of each perspective

(section 1.3.2.) and I will make comments and conclude in section 1.3.3.

42

1.3.1. Data

The general picture shows that, if around 1965 the level of fertility (TFR) in

Eastern European countries varied between 2 and 2.5 children per woman (Althus,

1992:140), after 1996 only Albania is above replacement level (Graphic 1.12). Gradually

or very fast, all other countries fell below replacement during this period of time despite

of various policies toward increasing fertility some Eastern European governments

adopted.

Fertility levels in Eastern European countries are similar to or lower than in

Western countries, although patterns of marriage are different (Sardon, J.P.:119): until

1990, for example, total marital fertility rates are around 28% lower in Western countries

than in Eastern Europe mainly as a result of a much higher percentage of consensual

unions in Western than in Eastern countries (Sardon: 137).

Graphic 1.1210. TFR during 1960-1998, selected Eastern European countries

3.5

3Slovak Republic Romania Poland 2.5 Bulgaria 2 Czech Republic Hungary 1.5

1

0.5

0 1960 1965 1970 1975 1980 1985 1990 1995 1998

10 I did not include Albania in this graphic because data are missing for most of the years. However, it is known that demographic transition began very late in Albania in comparison with other Eastern European countries (Falkingham&Gjonca, 2002). In 1960, in Albania, TFR was 6.85 and in 1996, 2.7 (Recent Demographic Developments in Europe, 1999) 43

source: Recent demographic developments in Europe, 1999

How is it possible to have a high and almost universal pattern of marriage and a low rate of fertility? For the Balkan countries the answer is: using abortion. As can be learned from

Table 1.10 and 1.11, the rate of abortion in these countries is around 1 or higher than 1 and much higher than in other countries. Romania has a strong tradition of using abortion; the highest abortion rate ever recorded in statistics was in Romania in 1965 – 252 per 1000 women (Henshaw&all:S33).

Table 1.10. Rate of abortion Table 1.11. Number of abortions, 1995

Region and No.of abortions (millions) Country and year Abortions Rate subregion Total Legal Illegal Bulgaria (1988) 111700 0.951 Europe 7.7 6.8 0.9 Yugoslavia (1987) 365700 1.018 Romania (1990) 992300 3.153 Eastern Europe 6.2 5.4 0.8 Czechoslovakia (1987) 156000 0.729 Northern Europe 0.4 0.3 - Hungary(1987) 84500 0.671 Southern Europe 0.8 0.7 0.1 Poland (1987) 122600 0.202 Western Europe 0.4 0.4 -

Source: F. Althus (1992) Source: H.Davis

After 1990, the mean age at first birth increased in most of the Eastern European

countries, which made some researchers conclude that the low fertility observed in the

region is actually a result of postponement of childbearing rather than a quantum

decrease. Bongaarts and Feeney’s (1998) influential article on the adjusted TFR was

enthusiastically incorporated into several articles trying to prove that the fertility

perspectives of Eastern Europe are rosier than the actual data predict. The authors

considered two components of the TFR: a quantum and a tempo component and the

method proposed by them tries to adjust the quantum component of the TFR by taking in

44 account the distortions introduced by the tempo component. Philipov and Kohler (1999) computed an adjusted TFR (using the Bongaarts-Feeney adjustment) for several Eastern

European countries and concluded that postponement of childbearing plays a certain role in the observed low fertility in Eastern Europe. However, several critics of the Bongaarts-

Feeney adjustment emphasized that “the method does not incorporate enough information to disentangle effects of fertility timing and level without imposing stringent and unrealistic restrictions” (Kim&Schoen: 2000) and that the method “is based on unsuitable fertility measures “ (von Imhoff & Keilman: 2000).

It is still unclear what role, if any, postponement plays in declining fertility in

Eastern Europe. On one hand, although for some Eastern European countries there were observed tempo changes (Philipov, Kohler, 1999), there are cases (Bulgaria, 1991, 1992) where the decrease in fertility was the sole result of quantum, not tempo changes, so it is unclear if women postpone childbearing or just give up having children.

On the other hand, the fact that fertility in Eastern Europe has been very low (1.5-

1.3) for more than a decade makes less credible the idea that women will catch up soon after postponing childbearing and a sudden increase in the number of children will be recorded. Longitudinal measures (as completed cohort fertility) are not affected by the postponement of childbearing, but they can be recorded only after the cohorts leave the childbearing ages (over 45). The fertility of cohorts shows that higher order fertility decreased in the cohorts born between 1935 and 1950. On the other hand, the region is not homogeneous. Higher order fertility in Poland and Czechoslovakia is about double that in Bulgaria, East Germany and Hungary (Althus:141). In Graphic 1.13 I present data on the fertility of cohorts of females born during 1945-1965. Although this is not

45

completed cohort fertility for all generations, it gives some perspective on how fertility

evolves.

Graphic 1.13. Cohort fertility for females born during 1945-1965 in Eastern European countries

2.6 Slovak Romania 2.4 Republic 2.2 Poland Czech2 Republic Hungary 1.8 Bulgaria 1.6 1.4 1.2 1 1945 1950 1955 1960 1961 1962 1963 1964 1965

source: Recent demographic developments in Europe, 1999

46

1.3.2. Theories

Scholars have attempted to explain the low fertility levels in Eastern Europe from various perspectives. Three of these perspectives have become well known and were reinforced on various occasions: one focuses on poverty as the main factor, the other on anomy and a third one considers the second demographic transition as explaining the fall of fertility.

a) Economic explanation

After 1990, as a result of political and economic changes in the region, many people fell into poverty and continued to remain poor for the rest of the transition period.

Inflation manifested all over, many countries imposed a monetary reform in an effort to stabilize their currencies and to help national economies. GDP fell in all countries of the region during the first part of the 1990s (Table 1.12), and although some of the countries achieved a good economic increase and improved the average standard of life, economic inequality exploded after 1990 and many people remained under the poverty line.

Table 1.12. GDP in Eastern European countries, 1989-1992 (1990 Geary-Khamis Dollars) Bulgaria Czechoslovakia Hungary Poland Romania USSR Yugoslavia 1988 6336 8675 6929 5790 4035 7032 6026 1989 6217 8729 6787 5685 3890 7078 5917 1990 5764 8464 6348 5113 3460 6871 5458 1991 4516 7244 6010 4798 2887 5793 4861 1992 4054 6845 5638 4726 2565 4671 3887

source: Maddison, 1995

47

Researchers using the economic approach to explain fertility decline in Eastern

Europe argue that raising a child is increasingly costly nowadays and, therefore, many

couples prefer to limit their fertility. At the same time, childlessness is not an option for

the majority of women, and this feature distinguishes Eastern from Western countries11.

In an article focused on fertility regulation in Bulgaria, Carlson & Omori (1998) note that the declining state socialist economy, followed by ‘rapidly changing socioeconomic conditions’ (Carlson&Omori:187) after 1990 made more and more women use both contraception and abortion as methods for limiting their fertility. Macura (2000), in research dealing with poverty and fertility in Eastern and Central Europe, argues that the economic problems of transition from a Communist economic system to an open market had a significant effect on falling fertility.

b) Uncertainty about the future

Although the idea of a relationship between income uncertainty, as a result of economic and social changes that Eastern European countries faces, and low fertility appears in several studies and seems to make a lot of sense (Philipov, 2001), the evidence to support such a relationship is very scarce. Ranjan (1999), using a two-period stochastic model of fertility behavior, argues that it is economically rational for individuals to postpone childbearing during the times of economic uncertainty. He concludes that ‘the increase in uncertainty about future income… gives rise to a threshold behavior such that

11 For example, in 1999, among women age 35-39, only 7.7 did not give birth to at least one child (Reproductive health Survey, 1999:29) 48 individuals above the threshold level of income want to have a child immediately, while those below the threshold wait for the uncertainty to be resolved’(Ranjan:17).Although

Ranjan suggests that his results can be used to explain fertility decline in Eastern Europe after 1990, he does not explain how to reconcile his conclusions with the fact that poor people in Eastern Europe, who obviously are more exposed to income uncertainty, tend to have more children compared to people with higher income (Reproductive Health

Survey, Romania, 1999:32).

c) Second demographic transition

The theory of the second demographic transition is nowadays one of the most popular explanations for the demographic changes occurring in Western Europe after

1970. First advanced by Lesthanghe (1983) as an explanation for contermporary trends in fertility and marriage in Europe, the idea of a second demographic transition seems to be a successful cultural explanation linking declining fertility, declining rates of marriage, increasing cohabitation and divorce to the ideational change that affects post modern

Europe. Desacralization of life and individualism make people less willing to raise a high number of children or to commit themselves to a lifetime union and, as a result, fertility and nuptiality rates decline. Initially developed at a macro level, this theory was later tested at a micro level and received a fair amount of support too. Although not many studies published in English or French argue for a second demographic transition explanation of fertility trends in Eastern Europe, the idea seems to be rather popular among Eastern European students (Philipov, 2001:13). However, studies reinforcing such

49

an explanation focus on Central rather than on South-Eastern Europe, and use similar

arguments for these countries as were used for Western European countries. Sobotka,

Zeman & Kantorová (2001), for example, argue that after 1990, demographic changes

characteristic of the second demographic emerged simultaneously in the Czech Republic

(Sobotka, Zeman & Kantorová:10), making it impossible to establish which one caused

the others. The authors conclude that, since the early 1990s, ‘there is no doubt that the

second demographic transition has been taking place in the Czech Republic’ (Sobotka,

Zeman & Kantorová: 15). Frątczak & Chmielewska (2001) explain that the new lifestyle, promoted through mass media and adopted by the young generations, makes people consider having children as a minor value. The decline in fertility in Poland, they say, increasing age at first marriage and the slow increase in cohabitation characterize the second demographic transition in Poland. The idea of the second demographic transition seems to be particularly popular in Poland and in 1999 a book called “Demographic

change in Poland during the 90s in the light of the second demographic transition” was

even published (Philipov, 2001:12).

d) Women’s status and fertility

The relationship between women’s status and fertility in Eastern Europe is not a

popular direction of study, as these two subjects are hardly studied together. However,

this seems to be a promising line of research. McDonald (2000) observed that, among the

low fertility countries, those countries which preserve strong inequality within the family

and family oriented institutions have the lowest fertility. He suggests that having a child

50 in such places is too heavy a burden for a woman and this is why many women decide to postpone or to avoid having children. On the other hand, students of women’s position in former Socialist countries (Funk, Nanette& M. Mueller, 1993, Gal & Kligman, 2000)) observe that, in spite of, or most probably because of, more than fifty years of strong political pressure for raising women’s status and a large participation of women in the labor force, the status of women is still low in Eastern European countries. This observation suggests that Eastern and Southern European low fertility could be the result of similar factors: the status of women within the family and family related institutions.

51

1.4. Conclusions

For most Eastern European countries, below replacement fertility was very much the case before the 1990s and what was happening after seems to be only a continuation of what began during the first part of the 1980s. Although it is true that fertility decreased in Eastern Europe while remaining at a constant level in Western Europe after 1990, this evolution could be seen as not necessarily a problem of Eastern Europe, but as a result, for example, of more friendly fertility policies in Western than in Eastern Europe after

1990. On one hand, there is no reason to believe that fertility must remain at a Western level in Eastern Europe. It can be argued that decreasing fertility is expected in Eastern

Europe unless the governments adopt policies toward stopping this trend. In fact, not long ago ‘East’ was considered completely different than ‘West’ in Europe, in terms of family formation characteristics, (Hajnal, 1982). Thus it is hard to understand why, ‘suddenly’, after 1990, it should be similar rather than different. On the other hand, ‘Western fertility style’ – defining as ‘Western’ all European countries that did not experience a state socialist system after World War II–is hardly as homogeneous as can be inferred from compounded statistics. Southern Europe has, for example, similar levels of fertility as

Eastern Europe, although history and socio - economic conditions of life are different for the two regions (Graphic 1.14).

52

Graphic 1.14: TFR in Western, Southern, Eastern and Northern Europe

3 2.8 2.6 2.4 Ea s te rn E. 2.2 Northe rn E. 2 Southern E. 1.8 We s te rn E. 1.6 1.4 1.2 1

5 0 5 0 5 0 5 0 5 0 -5 -6 -6 -7 -7 -8 -8 -9 -9 -0 0 5 0 5 0 5 0 5 0 5 5 5 6 6 7 7 8 8 9 9 9 9 9 9 9 9 9 9 9 9 1 1 1 1 1 1 1 1 1 1

source: UN (http://esa.un.org)

Therefore, is the fertility decline in Eastern Europe after 1990 a false problem? In a way, I believe it is. There is more and more evidence that Eastern Europe no longer exists as a coherent region, if it ever existed as one. Some of the countries of former

Eastern Europe joined the EU, so nowadays there is no political unity of the region. From an economic point of view, Eastern Europe was never homogeneous, with some of its countries (Hungary, Czech Republic) being considerably richer than others (Bulgaria,

Romania). Fifty years of communism were not able to homogenize populations that came from such different ethnic, cultural and historical backgrounds, and this can be easily seen from how different countries of the region evolved after 1990. Studies of ‘all

Eastern Europe’ seem to be popular only in demography. Other disciplines (such as

Political Science) use much smaller units, such as Central Europe or Balkan area, or comparisons of two countries or case studies of one country. There is no methodological reason to introduce all Eastern European countries in the same report, because the

53 economic and social conditions vary a lot nowadays in the region. A history of fifty years of communism does not unify the region anymore. Other factors play a role in the demographic evolutions of these countries. The future of comparative fertility studies in

Eastern Europe will focus, I think, more on comparisons between countries that really followed the same trend, which have common social, economic or cultural features, and will combine micro and macro analyses.

54

Chapter 2. The Politics of Fertility and Their Effects in Romania

2.0. Introduction

During the last century, Eastern European countries experienced various social, economic and political changes: the proportion of urban population as well as the educational level of people increased, the ethnic composition and boundaries of countries changed various times, and times of economic prosperity were followed by periods of extreme poverty and social anomy. All these transformations, together with social and population policies that various governments adopted, played an important role in the demographic characteristics of Eastern European populations. Because countries have various histories, it is difficult to explain the evolution of a certain demographic phenomenon without taking in account the history of that population.

In the following chapter, I will focus on the fertility evolution of one Eastern

European country, Romania, and I will show how economic, social and policy history contributed to the fertility trend. There are two general ideas for which I will argue in this part: 1) fertility evolutions of different social and ethnic groups vary, and the proportions of these groups in the general population have an important influence on fertility at national level; and 2) the higher than replacement fertility during the period of 1967-

1989 was not a natural phenomenon, but a direct result of the political restrictions on using contraceptives and performing abortions. If this is the case, the low fertility immediately after 1990 is not a “strange” phenomenon, but an expected trend, only

55 interrupted during 1967-1989. Although the fall of fertility after 1993 in Romania can be seen and partially explained by the social and economic changes following the fall of communism, the decline of fertility during 1989-1992 is also a result of the changes in abortion policies at the end of 1989. It is a classical case of unmet need: women had more than two children on average before 1989 just because they were not able to control properly the number of children.

In arguing for these ideas, I am going to use three types of data:

1) Census data, from 1966, 1977, 1992 and 2002 censuses;

2) Aggregate data from fertility surveys done in 1967, 1974 and 1978

3) Individual level data from a fertility survey done in 1999.

The aggregate level statistics, useful in giving a general frame for what happens with a certain phenomenon, do not allow one to link individual characteristics with the number of births. They present only the general picture (how many children were born in a certain year in rural or urban areas, how much education the mothers of children born have, etc.). The use of only aggregate level data in the study of a certain subject can lead to many problems, studied in the literature (ecological fallacy, for example, Firebaugh,

1978). This is why, in order to understand how fertility relates to other social and economic factors, in this chapter I will use both individual and aggregate level data.

56

2.1. History matters

2.1.1. Short political history of Romania

Situated in the Southeastern part of Europe, in Balkan Peninsula, Romania is a relative young state. Before the 19th century, on the current territory of Romania there were four different states: Moldavia, Walachia, Besserabia and Transylvania. Moldavia and Walachia, which for hundred of years were Turkish vassals, elected in 1859 the same domnitor (chief of state) and this is how the first state named Romania appeared on the

World map. In 1877, Romania declared independence from the Turkish (Ottoman)

Empire. It became a kingdom in 1881 with the first king being Carol I, from an Austrian family, closely related to king of Austro Hungarian Empire. In 1918, after World War I, and following the disintegration of the Austro-Hungarian Empire, Transylvania,

Besserabia, and a small part of Bulgaria (Cadrilater) joined Romania. After World War

II, USSR incorporated Besserabia, known now as Republic of Moldova, and Cadrilater became again part of Bulgaria.

From the point of view of ethnic composition, Romania has nowadays a relative homogenous population, with about 90% of the population declaring a Romanian ethnicity, 7% Hungarians and 2% Rroma (Gypsy). The ethnic composition of the population changed dramatically during the last 70 years, and that happened for various reasons (Table 2.1).

57

Table 2.1. Ethnic composition of Romania in 1930 and 2002

1930 2002

Romanians 71.9 89.5 Hungarians 7.9 6.6 Germans 4.1 0.3 Jews 4 * Ukrainians 3.2 0.3 Russians 2.3 0.2 Bulgarians 2 * Rroma (Gypsy) 1.5 2.5 Turks 0.9 0.1 Serbs/Croats/Slovens 0.3 0.1 Czech/Slovaks 0.3 0.1 Polish 0.3 * Greeks 0.1 * Tatars 0.1 0.1

source: Romanian Censuses 1930, 2002

Ukrainians, Russians, Greeks, Turks and Bulgarians lived mainly in regions that are no longer part of Romania; this is why their percentage in the total population decreased. Most of the Serbs (Croats, Slovenians), Czech and Polish probably assimilated into the general population and their children or grandchildren do not identify themselves anymore with these ethnic groups. The percentage and number of Jews and ethnic

Germans is lower in 2002 because of emigration: after World War II, governments of

Federal Republic of Germany and Israel negotiated with the Romanian government to let people with a German or Jewish origin emigrate. Many of the ethnic Germans and Jews took advantage of these treatises and out-migrated. The percentage of Hungarians decreased because of three different reasons: a) for a short time after World War II, the

58

Hungarian and Romanian governments had an exchange agreement, so Hungarian citizens with Romanian ethnicity were able to go and live in Romania while Romanian citizens with a Hungarian origin would go and live in Hungary; b) after 1990, many

Hungarians went to Hungary for work, because the salaries and living conditions are much better there than in Romania and c) Hungarians always had lower fertility than many other ethnic groups in Romania. Rroma (Gypsy) is the only minority group that increased both as a number and percentage during the last 70 years. The Gypsies appeared in Europe about 700 years ago, probably coming from the Punjab region, India

(Barany, 2002:9). Those living in Eastern Europe are Vlach Gypsy while in Western

Europe (Germany) lives a different group (Sinti). In Romania there are about twelve different Gypsy tribes (Aurari, Lingurari, Biesi, Caravlahi, etc), originally members of a large tribe called Rudari (Barany: 12). Gypsy women marry very young and used to have many children. This is one reason why the Gypsy population in Romania (as well as in other Eastern European states) increased considerably over the years. Although undercounting alone cannot explain the large change in Gypsy population, a second reason could be related to census errors: counting the Gypsy population is a problem anywhere, so the figures should be used cautiously.

59

2.1.2. Communism, abortion and eugenics

The demographic transition began in most of the Romanian provinces at the beginning of the 20th century, when abortion began to be used extensively as a method of limiting the number of children (Graphic 2.1). As the newspapers of these times show

(Bucur, 2002), around 1920, abortion began to get more attention from the media within the more general framework of eugenics measures. Eugenic ideas played an important role in Germany as well as in other European countries in the 20th century, and they were incorporated into the Nazi ideology.

Graphic 2.1. Rate of natality, Romania, 1899-1966

50 45 40 35 30 25 20 15 10 5 0

5 1 9 0 5 99 02 0 08 1 14 23 26 2 32 38 47 5 53 59 62 6 9 9 9 9 9 18 1 19 1 19 19 1920 19 1 19 19 1935 19 1 19 19 1956 19 1 19

sources: Romanian Yearbook, 1935-1936, Ghetzau (1997), Romanian Demographic Yearbook (2001)

In Romania, around 1920, several articles in some of the most important newspapers

(Adevarul, Universul) discussed, from an eugenic point of view, whether abortion should

60 or should not be legalized and regulated by the state, taking into account that many women, especially from urban settings, use abortion to control their fertility (Bucur:201).

In 1936, Sabin Manuila, an important public figure (statistician and demographer) published an article “in which he emphatically stressed the need to negotiate a modus vivendi between the state’s concern for the growth of the birth rate and its responsibilities in protecting the health and rights of individual citizens” (Bucur:204).

In 1936, the parliament adopted a law which criminalized abortion in all cases with the exception of those couples in which one or both parents were mentally ill (Bucur: 206).

In 1948, after the Communist Party assumed power, this law (Article 482, Romanian

Penal code) was revised, but abortion remained a criminal activity (Kligman:47). During the communist regime, population policies had a strong ideological component: having children was considered a duty to the nation, a biological weapon against the capitalist enemies. After Stalin’s death (in 1953) the ideological direction changed, emphasizing women’s rights as citizens to be equal with men, and in 1955 the USSR liberalized abortion (Avdeev, 2001:13). The other state socialist countries followed the new direction: Poland liberalized abortion in 1956 (Fratczak, 2003), Czechoslovakia in 1957

(Sobotka & all:4), Bulgaria in 1956 (Vassiliev, 1999). In September 1957, Romania adopted the law, which made abortion legal (Decree number 463 for legalizing interruption of pregnancy, published in Buletinul Oficial nr.26/30, September 1957). By the end of the 1950’s, Romania began a period of important economic changes: fertile land became state’s property, there was a high rate of urbanization and industrialization, and there were an increasing levels of education and female labor force participation

(Graphic 2.2).

61

Graphic 2.2. Percentage of urban population, Romania, 1930-2000

60

50

40

30

20

10

0

0 5 0 5 0 5 0 5 0 5 0 5 0 5 0 3 3 4 4 5 5 6 6 7 7 8 8 9 9 0 19 19 19 19 19 19 19 19 19 19 19 19 19 19 20

source: Romanian Demographic Yearbook, 2001

All these changes affected the rate of fertility, and by 1966, the average number of children per women (TFR) fell below replacement level. The communist authorities began to be concerned over the decline of fertility, especially its effects on the future workforce (Bucur, 227) and interpreted this trend as a result of legalizing abortion in

1957.

Various European governments adopted policies attempting, explicitly or implicitly, to raise the level of fertility during the last century. Among these policies, the prohibition of abortion adopted by the Romanian government in the end of 1966 was seen as the most drastic measure, with many negative consequences. A huge increase in fertility followed during the first three years after this policy was adopted (Graphic 2.3), although in the long term, the number of births continued to fall. This law and its effects were largely discussed in articles and books at that time, but also recently. David &

Wright (1971) discussed the effects of abortion prohibition in Romania on maternal and infant mortality. Ghetzau (1971) and Teitelbaum (1974) studied which age categories of

62 women were more responsive to the policy. In 1998, Kligman (Kligman, 1998) published a book evaluating the negative effects of the policy on the day-to-day life of women.

Graphic 2.3. TFR in Romania, 1956-2000

4.0

3.5

3.0

2.5

2.0 TFR

1.5

1.0

0.5

0.0

1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

source: Romanian Demographic Yearbook, 2001

The law was carefully prepared by a mass media campaign. Beginning in August

1966, Scanteia, the Romanian Communist Party newspaper, began to publish articles dealing with maternity as a woman’s duty. Usually Scanteia had the first two pages full of political news: foreign authorities visiting Romania, Romanian authorities going abroad, new economic success of the Romanian Communist Party, etc. However, after

July 1966, things changed, and the first two pages began to include pronatalist articles.

For example, on August 4, 1966, Ursuala Schiopu, a psychologist, argues that “feminine charm is regenerated by maternity” (Scanteia, 08/04/1966, page 2), on August 17, on the first page is an article on the schools and children (Scanteia, 08/17/1966, page 1). On

August 21 there was an article about “Children’s love” (Scanteia, 08/21/1966, page 2),

63

September 19 – in an article called “Maternity’s advantages”, a physician argues that

pregnancy heals acne, eczemas, depression, and melancholy (Scanteia, 09/19/1966, page

2). On September 23, Ilea states that women ask for abortion only because they are superficial and it is easy to get an abortion (Scanteia, 09/23/1966, page 2 ), and a day later, Zaharia Stancu, an important Romanian writer and academician, discuss about past times when families had many children and all of them were happy (Scanteia,

09/24/1966, page 1). On September 25 an article called “Maternity” presents various stories about abortions, concluding that ‘very few women ask for abortion because of objective reasons’ (Scanteia, 09/25/1966, page 1-2).

Finally, on October 2, 1966, the new decree (770/1966) called “Measures to regulate pregnancy interruption” was published on the first page. The decree, it was argued, was designed as a ‘way to help families with many children, to increase natality and to improve the care for mother and child” (Scanteia, 10/02/1966, page 1). The new decree stated that abortion was forbidden unless one of the following was true: pregnancy was a danger for woman’s life, one parent had a transmissible genetic disease, a woman was 45 or older, she gave birth and has in her care four or more children, or the pregnancy is a result of incest or rape. Although fertility increased dramatically after this law was adopted, some other unwanted consequences occurred: infant mortality and maternal mortality raised dramatically (Graphic 2.4). For the next three months, Scanteia

and some journals for women continued to publish numerous articles design to educate

the public toward having more children.

64

Graphic 2.4. Infant mortality, 1960-1972

80 70 60 50 40 30 20 10 0

0 1 4 5 6 7 8 9 2 6 6 7 196 196 1962 1963 19 19 196 196 196 196 1970 1971 19 .

source: Romanian Demographic Yearbook, 2001

Interestingly, as Bucur (2002) observed, “the language of this law [Decree 770/1

October 1966] … echoes that of the Carol II Penal Code [from 1938], which had been passed with eugenic concerns in mind: “ Interrupting pregnancy represents an act with grave consequences for the health of the nation and brings serious prejudices to the nation’s population growth” (Bucur:228). It seems that the eugenics ideas, very popular at the beginning of the 20th century in Romania, but totally repudiated by the communists as reminiscent of Nazi ideology, resurfaced with Decree 770. In fact, at the beginning of the 1960s, Ramneantu, one of the supporters of the eugenics movement in Romania, sent a number of memoranda to the Central Committee and Ministry of Health asking for

“controlling women’s access to birth control and fostering more ‘responsible’(i.e. , eugenically minded) behavior among couples of childbearing age” (Bucur:228).

Although it is hard to estimate how much importance the authorities gave to his memoranda, it is interesting that the language and arguments used by Ramneantu appeared in the body of the law.

65

In the end of 1989, as in many other Eastern European countries, the communist government lost power in Romania and the new government abolished the abortion law.

In the following year (1990), the total fertility rate fell from 2.2 (1989) to 1.8 (1990), and there were estimated to be 3 abortions for 1 live birth (Romanian Ministry of Health).

After 1990, fertility continued to fall and by 2001, the TFR was 1.3, one of the lowest in

Europe.

66

2.2. Data and interpretations

2.2.1. Censuses

During the last forty years, four censuses were done in Romania: in 1966, 1977,

1992 and 2002. Data from these surveys regarding fertility show that, for all cohorts, fertility fell between 1977 and 2002, but for certain cohorts fertility was higher in 1977 than in 1966, no doubt as a result of the 1966 abortion law (Table 2.2).

Table 2.2. Number of children born per 1000 women, by age groups

1966 1977 1992 2002 15-19 77.8 99.7 67 53.4 20-24 683.9 789 646.8 398.6 25-29 1329.7 1670.8 1513.1 967.2 30-34 1792.7 2165.7 1967.4 1430.7 35-39 2159.1 2315.1 2140.3 1878.8 40-44 2464.6 2317.6 2286.8 2105 45-49 2728.2 2296.3 2343.4 2175 50-54 3071.3 2407.1 2281.4 2278 55-59 3328.8 2579.8 2197 2320 60-64 3559.1 2865.7 2159.4 2257 65-69 3656.6 3080.4 2260.7 2166 70-74 3834.7 3240.2 2411.4 2124.6

source: Romanian Censuses, 1966, 1977, 1992, 2002

The 1966 law against abortion had a remarkable influence on fertility in the short term, and changed significantly fertility in the long term. In 1967, the TFR was 3.7 children per woman, while one year before (1966) it was only 1.9. Teitelbaum (1972) showed that, for many age groups, the peak increase in 1967 (200% and more) in parity

67

specific birth rates were recorded for women who had two children (and gave birth to the

third child). Although after 1969, the TFR decreased, it was still above replacement level

in 1989 (2.2).

In order to evaluate the long term effects of the 1966 law, I used data from the

1966 census on the number of children women age 15 and over had, and I considered the

1966 parity specific birth rates as remaining constant over the next 25 years and mortality

rates as being 0. Then, I calculated the cohort fertility rates of cohorts born in 1922-

194612 (those who were age 20-45 in 1966) resulting from this simulation and I compared

them with the real cohort fertility rates of these cohorts, as shown from 1977 and 1992

censuses (Table 2.3, 2.4).

Table 2.3. CFR from simulation and from 1977, 1992 censuses

Simulation 1977* 1992* 1942-1946 1.21 2.11 1937-1941 1.60 2.34 1932-1936 1.89 2.32 2.28 1927-1931 2.14 2.30 2.20 1922-1926 2.39 2.41 2.16

source: Censuses 1977, 1992 and author’s computation

Table 2.4. Percentage of childless women

Simulation 1977* 1992* 1942-1946 26% 11% 1937-1941 16% 12% 1932-1936 14% 10% 13% 1927-1931 14% 13% 16% 1922-1926 16% 16% 19%

source: Censuses 1977, 1992 and author’s computations

12 Parity specific fertility rates were only available for five year age groups. I did not include the age group 15-19 in the computations because they were only 40-44 in 1992, so their fertility was not completed. 68

* because the 1977 census was taken 11 years after the 1966 census, there is a 1 year difference in the year of birth between the simulated cohorts and those recorded in 1977. I assumed that this difference did not play a major role in the results. The same is true for 1992 census, which is 26 years, instead of 25 years, later than 1966 census.

The results show that the effects of the 1966 law were significant for cohorts born in 1932-1946 (women age 20-34 in 1966) while the other cohorts (those over age 34 in

1966) did not record much of a change in their fertility because of the 1966 law. With the

1966 age groups parity specific fertility rates, fertility would be way below the replacement level for cohorts born in 1942-1946, so the drop of fertility below replacement would have occurred well before 1989.

Fertility is influenced by various factors, such as education, nationality, residence

(urban, rural), and changes in general fertility are a result of: 1) changes in the fertility of various groups that form a population; and 2) changes in the proportion of the groups in the population. If we compare the cohort fertility13 of women age 40 and over in 1977,

1992, 2002, for various levels of education, it can be seen that fertility did not change much (for some cohorts it actually increased), but the composition of population changed remarkably between 1977 and 2002, so a good proportion of fertility change is actually due to this change in composition.

Table 2.5. CFR for college educated women, 1977, 1992, 2002

1977 1992 2002 40-44 1.20 1.42 1.26 45-49 1.07 1.40 1.33 50-54 1.00 1.29 1.42 55-59 0.97 1.18 1.39 60-64 0.93 1.06 1.28 65-69 0.86 1.00 1.18 70-74 0.82 0.98 1.03

13 Women age 40 and over have very low fertility rates, so I assumed that fertility is completed at age 40. 69

source: Censuses, 1977, 1992, 2002

Table 2.6. CFR for high school educated women, 1977, 1992, 2002

1977 1992 2002

40-44 1.42 1.70 1.70 45-49 1.29 1.61 1.67 50-54 1.19 1.48 1.72 55-59 1.18 1.38 1.62 60-64 1.23 1.26 1.48 65-69 1.30 1.17 1.37 70-74 1.25 1.17 1.24

source: Censuses, 1977, 1992, 2002

Table 2.7. CFR for elementary school educated women, 1977, 1992, 2002

1977 1992 2002 40-44 2.61 3.15 3.37 45-49 2.57 2.94 3.25 50-54 2.66 2.70 3.07 55-59 2.84 2.46 2.87 60-64 3.14 2.38 2.65 65-69 3.35 2.44 2.40 70-74 3.49 2.61 2.32

source: Censuses, 1977, 1992, 2002

Table 2.8. Women with a college degree for all age groups, 1977, 1992, 2002

college 1977 1992 2002 20-24 1.6% 2% 5% 25-29 7.9% 8% 14% 30-34 6.9% 9% 11% 35-39 3.9% 8% 10% 40-44 2.8% 8% 10% 45-49 2.4% 7% 9% 50-54 1.7% 4% 9% 55-59 1.5% 3% 7% 60-64 1.6% 3% 4% 65-69 1.9% 2% 3% 70-74 1.2% 2% 3% 70 source: Censuses, 1977, 1992, 2002

Table 9. Women high school educated for all age groups, 1977, 1992, 2002

1977 1992 2002 15-19 6% 14% 11% 20-24 26% 51% 49% 25-29 16% 44% 35% 30-34 8% 33% 42% 35-39 7% 26% 42% 40-44 6% 22% 34% 45-49 5% 14% 25% 50-54 4% 12% 30% 55-59 4% 10% 12% 60-64 4% 8% 11% 65-69 5% 6% 10% 70-74 3% 6% 8% source: Censuses, 1977, 1992, 2002

Table 2.10. Women with elementary school education, 1977, 1992, 2002

1977 1992 2002 15-19 11.7% 6% 12% 20-24 11.1% 2% 5% 25-29 20.0% 2% 3% 30-34 50.2% 4% 2% 35-39 58.9% 6% 3% 40-44 66.6% 12% 4% 45-49 61.5% 33% 7% 50-54 66.2% 42% 13% 55-59 72.8% 47% 33% 60-64 74.8% 44% 41% 65-69 78.3% 46% 44% 70-74 84.8% 52% 42% source: Censuses, 1977, 1992, 2002

71

When comparing 1992 with 2002, college educated women are those who reduced their number of children most, although in comparison with 1977, college educated women are not much different in 2002 than in 1977 in terms of fertility. Fertility changes are significantly influenced by the changes in women’s level of education: cohort fertility rates for women with a similar level of education and in a similar age group did not decrease significantly between 1977, 1992 and 2002, but the distribution of women on various levels of education changed. If we use the same structure of education for 2002 as in 1992, and we keep the same CFR as in 2002 for each education level group, then 75% of the difference between the cohort that was 45-49 years old in 1992 and the cohort that was 45-49 years old in 2002 is explained by changes in the education distribution and only 25% is explained by changes in the fertility rates. In 2002, cohort fertility rates by education levels are actually higher than in 1977, so the changes in the cohort fertility rates (for the cohort as an whole) is only the result of changes in the education distribution. Education was not the only factor by which fertility varies; the ethnic composition is also important. Romanians, Hungarians, Germans and Gypsy were the most numerous ethnic groups in Romania after 1966 (Table 2.11), and they had different levels of fertility.

72

Table 2. 11. CFR for Germans, Hungarian, Romanian and Gypsy women in 1992, 2002

Romanians Hungarians Gypsy Germans age group 1992 2002 1992 2002 1992 2002 1992 2002 40-44 2.3 2.1 2.2 2.0 4.0 4.0 1.9 1.8 45-49 2.3 2.1 2.1 2.1 3.9 4.1 2.0 1.8 50-54 2.3 2.3 2.1 2.2 3.7 4.0 1.9 1.8 55-59 2.2 2.3 2.0 2.1 3.6 3.9 1.7 1.9 60-64 2.2 2.3 2.0 2.0 3.5 3.7 1.6 1.7 65-69 2.3 2.2 2.0 2.0 3.6 3.7 1.5 1.7 70-74 2.4 2.1 2.2 1.9 3.6 3.5 1.5 1.5

Among these groups, Germans, Hungarians14 and Romanians had similar fertility, while Gypsy fertility is higher. Some authors (Kligman, 1998) suggest that, alarmed by the high fertility of the Gypsy population, the Romanian government implemented differently the antiabortion legislation, being more permissive toward abortions done by

Gypsy women. The only measure that could be interpreted as limiting Gypsy fertility was that the government made child allocations conditional “upon employment of one of the parents and school attendance by the children” (Bucur, 2002).However, there is no serious evidence that this measure had any effect of Gypsy fertility, and as the cohort fertility rates show in 1992, the Gypsy had two times more children than the other ethnic groups had. The 1966 pronatalist law included some increases in family allowances, which were independent of family income, and, as Gathier and Halzius (1997) observed, family benefits that are independent of family income (meaning that they are not calculated as a percentage of the family income) have a higher impact on the fertility of low income families, because the cost of children is lower for low income families than for high-income ones.

14 In many cases, Germans and Hungarians have lower fertility than Romanians. 73

Changes in the ethnic composition of population between 1992 and 2002

contributed positively to the fertility. I used the five largest ethnic groups (Romanians,

Hungarians, Germans, Gypsy and Ukrainians), which account for 99% of population, and

data regarding fertility and proportion in the population for different age groups of these

ethnic groups. If we assume that the ethnic composition is the same in 1992 and 2002,

and we keep the number of children born by various age and ethnic group combined as it

was in 2002, then the fertility of the general population would be lower than it was

recorded in 2002 (Table 2.12).

Table 2.12. Number of children born per 1000 women in 2002, and simulated with 1992 ethnic structure

simulated 2002 15-19 46.97 52.73 20-24 387.79 393.83 25-29 946.54 956.09 30-34 1409.70 1416.00 35-39 1845.65 1854.32 40-44 2071.49 2079.82 45-49 2143.02 2151.31 50-54 2244.89 2252.44 55-59 2291.72 2293.05 60-64 2228.18 2230.62 65-69 2133.34 2139.90 70-74 2091.58 2098.46

This happens because Hungarian and German ethnics, with low fertility, have

lower percentages in the population in 2002 in comparison with 1992, while Gypsy and

Ukrainians, with higher fertility, have a higher share of population in 2002, at all age

groups, but especially in the young age groups. If the fertility of ethnic groups does not

change very much in the future, due to these changes in the structure of population,

74 fertility will be higher in the next years, with an increasing share of children being born in Gypsy and Ukrainian families, and fewer into Hungarian and German families.

75

2.2.2. Fertility surveys: 1967, 1974, and 1978

After 1966, a number of fertility surveys were conducted by several institutions

(Institute for Public Health and Hygiene, Centre for Health Statistics, Central Department of Statistics and Ministry of Health) to estimate better the effects of the 1966 abortion law and to detect the main factors which played a role in keeping fertility below governmental expectations. These surveys, involving large female samples (over 10,000) and conducted by well trained teams of researchers, showed that there were two important reasons for which women refused to have many children. Firstly, female work force participation and an increasing level of urbanization left families without any help in taking care of children. Secondly, the attitude regarding large families changed: women preferred to have more comfort, and a high number of children would prevent them from getting a better life. In the following, I will summarize the main results of the surveys, focusing only on those results relevant for this study.

In 1967 the Institute for Public Health and Hygiene, Centre for Health Statistics,

Central Department of Statistics and Ministry of Health, decided to organize a large survey to investigate the factors that influence fertility in Romania. The sample included more than 10000 women at first marriage, age 15-49, chosen from regions with high, medium and low fertility. The survey was first done in 1967 and then repeated in 1974 and 1977 on a similar sample. These surveys were supposed to highlight the main factors responsible for keeping fertility low.

76

The sample consisted of 12,783 women with an age and residence distribution similar to that documented in the 1966 census (Ilea&all:26). On average, women considered that 2.5 children is the optimum for a family, with small differences between urban and rural areas: urban women thought that 2.38 is optimum, while rural women said 2.61. Even when the education level of women was taken in account, all groups said that a family has to have more than two children (women with college education, for example, considered that, on average, 2.15 children is optimum for a family). However, when the respondents were asked about the number of children they wished for at the beginning of marriage, most women wanted, on average, fewer than 2 children, and only the 20-29 age group had an average of 2.03 (Ilea&all:46). The actual number of children born was around 2: women age 40-49 had, on average 2.12 children, while women age

30-39 had 1.91 children. Those married for 5-9 years had 1.54 children; those married for

10-19 years had 1.98 children while those married for more than 20 years had, on average, 2.31 children. College educated women had, on average 0.98 children, women with elementary school education had 1.75 children while those without education 2.05.

Women age 15-19 planned to have, in the future, 1.42 children, which mean that they expected to have around 1.9 children during their life. That shows that as early as 1967, those cohorts that were 15-19 planned to have fewer than 2 children on average.

When asked why they want to have only one or two children, women age 30-49 said their work does not allow them to have many children (around 18% ), they married too late to have more than 1-2 children (around 15%), their house is unfit for having more children ( around 8%) or that life with few children is easier (~7%). The percentage of women which gave the last reason is almost double for younger women than for older

77 ones (9.3% for women 30-39, 5.1 % for women age 40-49). About 13% of women with a college education mentioned the last answer for having few children, while only 5% of women with elementary school used it as a reason. When all women who said they want only 1-2 children were asked why they do not want to have more children, 28% of those married for less than 4 years answered life with few children is easier. All these results can be interpreted as first signs of a second demographic transition which probably was delayed in Romania because of the government policies encouraging high fertility and discouraging divorce.

A new fertility survey was initiated in 1974, using a similar sample and questionnaire as in 1967. The sample included 11,078 women, at first marriage, age 15-

49.At the time of this survey, women considered the optimum number of children for a family to be 2.6, although, at the beginning of marriage, they wanted only 2.2 children.

On average, women age 40-49 in 1974 had 2.3 children: those living in urban areas had

1.8 while those living in rural areas had 2.7. Women without education had, on average,

2.5 children while women with college education15, 1.2. About 13% of women age 30-49, childless or with only 1-2 children, said that the main the main reason that they had few children was health problems, while around 10% of women said there is nobody to take care of the child when she goes to work (Muresan&all:36). The change in motivations for why women have two children or less (between the first survey (1967) and the second one (1974)) is interesting: a significant proportion of women said sickness is the reason for which they do not have many children, while in 1967 this was not cited as a reason. It is hard to believe that in 1974 there were more sick women than in 1967, taking into

15 For all age groups. 78 account that life expectancy increased between 1967 and 1974. The answers reflect an adaptation of women to the political conditions of the moment: in order to ask for a legal abortion, a woman had to prove she was sick, so many women had in their files that they suffer from various sicknesses that made them unfit for having more children.

The third fertility survey was done in 1978, on a sample of 10,141 women at first marriage, age 15-49, with an age distribution similar to that from 1977 census. Women age 40-44 living in urban areas had, on average, 1.89 children, while those living in rural areas had 2.92. During the first ten years of marriage, women had on average 1.99 children (1.83 in urban areas and 2.14 in rural areas) (CND: 15). Women with a college education, married for 15-19 years, had on average, 1.67 children, while women with elementary education, married for 15-19 years, had 2.8 children. After 10 years of marriage, women with high school had fewer than 2 children (1.8-1.5), and women with college degrees even less (1.6-1.4). On average, a woman had 4.5 pregnancies, 2.08 were unwanted and 2.2 abortions. Those women who lived in rural areas until they were 12 years old, and lived in urban areas in 1978 had, on average, 1.8 children, while the average for all women living in urban areas is 1.72. Those women who lived all their life in small towns had on average 1.56 children, those who lived in large towns hade 1.58 children.(CND:17).

79

2.2.3. Individual level data

This analysis was based on the Reproductive Health Survey Romania, 1999, a

survey conducted in 1999 by the CDC and Romanian Ministry of Health. The sample for

this survey (national representative) includes 6,888 women, aged 15-44 years who lived

in Romania at the time of the interview (1999). The questionnaire has 9 modules:

background characteristics, sex education, fertility/pregnancy, family planning

knowledge/sexual experience, current and past contraceptive use, women’s health,

reproductive health knowledge/attitudes, socioeconomic characteristics, aids/std and

violence. This research uses data from module 1 (background characteristics), module 3

(pregnancy/fertility history) and module 8 (socioeconomic characteristics). In Module 3,

information on the last 23 pregnancies was recorded: how and when these pregnancies

ended, if the child is still alive (in the case that the pregnancy resulted in a live birth) and

how old the child was when he died (for live births which later died).

As a method I used simple life tables and life tables with multiple decrements

(Schoen, 1988) for those women who were pregnant during times when various fertility

policies were implemented in Romania. For each period of time, I restricted the analysis

to women age 15-44 who were at risk of having a pregnancy (for example, those who

already had a pregnancy entered the sample for the second pregnancy, but those without

any pregnancy did not enter the sample for the second pregnancy).

Generally speaking, there are four main periods relevant for this study16:

16 See also Kligman, 1998:59. During 1957-1966, abortion was legal and contraceptives available, but the women included in this sample were not at reproductive age during that period of time 80

- 1966 -197317: abortion was banned for all women younger than 45, with the

exception of: 1) those pregnancies that resulted from rape or incest, 2) if a woman

already had four children in her care, 3) if a woman had some important physical

disabilities, 4) if the pregnancy was a danger for the woman’s life or 5) if a woman or

her male partner had a hereditary illness with serious consequences18. Although

contraceptives were not banned, they were no longer produced or imported;

- 1974-1984: new instructions regarding implementation of Degree 770 were emitted in

1974, in an attempt to control better abortions performed legally and illegally

- 1985 - 1989: a new decree19 increased the number of children and the age a woman

has to have in order to get a legal abortion (5 children, and 45 years);

- 1990 - : abortion became legal.

Because of the small number of cases for the 1966-1974 period of time, for some

of the analyses, I considered only three main periods: 1966-1984, 1985-1989 and 1990

and over.

This research uses two main models: one in which the rate of transition from the

state ‘not pregnant’ to the state ‘pregnant’ is computed and compared for periods of time

when various fertility policies were in force (Figure 2.5). The second model (Figure 2.6)

estimates the rates of transition from the state ‘non pregnant’ to ‘has a live (or still) birth’

and from state ‘not pregnant’ to ‘has an abortion’. Live births and stillbirths were

grouped together, because I assume a woman who had a stillbirth did not want to

terminate the pregnancy.

17 In 1972, Decree 770 was changed, and the new law permitted women age over 40 to have a legal abortion (before only women age 45 and over could ask for a legal abortion). The sample of this study does not include women which, at that time, were 40 or over, so this change is not relevant for this study. 18 Decree 770, October 1966 19 Decree 411, December, 1985 81

Figure 2.5. States for model 1 Figure 2.6. States for model 2

Respondent has Not a live (or still) not pregnant pregnant pregnant birth

Respondent has an abortion (induced or not)

Results

The analysis of pregnancy history shows that the majority of the first and second pregnancies end in a birth, for the whole sample and for each period of time (see Table

2.13-2.17). From 1974 to 1999, the proportion of pregnancies (of any order) ending in an abortion increased, showing that what happened after 1990 is not a strange evolution, but a continuation of what was before (although at a higher level).

Table 2.13. Distribution of births/abortions for first to fourth pregnancy

first second third fourth Result pregnancy pregnancy pregnancy pregnancy birth 81% 62% 43% 32% abortion 19% 38% 57% 68% Total 4877 3866 2752 1930

82

Table 2.14. Distribution of births/abortions for the first pregnancy

Years birth abortion Total 1966-1973 89% 11% 108 1974-1984 89% 11% 1707 1985-1989 85% 15% 1188 1990-1999 70% 30% 1874

Table 2.15. Distribution of births/abortions for the second pregnancy

Years birth abortion Total 1966-1973 70% 30% 30* 1974-1984 73% 27% 1132 1985-1989 71% 29% 1015 1990-1999 48% 52% 1689

Table 2.16. Distribution of births/abortions for the third pregnancy

Years birth abortion Total 1966-1973 67% 33% 3* 1974-1984 60% 40% 634 1985-1989 54% 46% 702 1990-1999 31% 69% 1413

Table 2.17. Distribution of births/abortions for the fourth pregnancy

Years birth abortion Total 1966-1973 100% 0% 1* 1974-1984 50% 50% 351 1985-1989 41% 59% 458 1990-1999 23% 77% 1120

Although the proportion of pregnancies resulting in abortion is higher after 1990, taking into account that before 1990 abortion was banned and that having an abortion was a crime, the difference between what happened after and before 1990 is actually not striking.

83

Life tables with multiple decrements for the first pregnancy (ended with a birth or an abortion) show that, by the age of 30, 83% of women age 15 and over during 1966-

1983 would have a first birth, 11% would have a first pregnancy ending with an abortion and about 6% would not get pregnant. During 1985-1989, 79% of women age 15 and over would have a first pregnancy ending with a birth by the age of 30, 14% would have an abortion and 6% would not get pregnant.

After 1990, 58% of women age 15 and over would have a birth by the age of 30,

26% would have an abortion and 16% would not get pregnant. As can be observed, the difference between the period after 1990 and the previous periods is mainly observed in the percentage of women not getting pregnant by the age of 30 (16% in comparison with

6% for the periods before). This shows that availability of contraceptives after 1990 had an influence on the timing of fertility in Romania.

Graphics 2.7 and 2.8 show the evolution of the rate of transition to a first birth and abortion during 1966-1984, 1985-1989, and 1990-1999. The major change that can be observed is that the rate for a first abortion is much higher for young women after 1990 than before (Graphic 6) as a result, probably, of a younger age at sexual intercourse after

1990.

84

Graphic 2.7..Rate of transition to a first pregnancy ending with a birth

0.250

0.200

0.150 1966-1984 1985-1989 0.100 1990-1999

0.050

0.000 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29

Graphic 2.8. Rate of transition to a first pregnancy ending with an abortion

0.060

0.050

0.040 1966-1984 0.030 1985-1989 1990-1999 0.020

0.010

0.000 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

The life table with multiple decrements for the second pregnancy shows that, during

1966-1984, 80% of women age 15-30 who were at risk20 experienced a second pregnancy

resulting in a birth, 15% had their second pregnancy ending with an abortion and 5% did

not experience a second pregnancy. During 1985-1989, 79% of women at risk age 15-30,

20 By ‘at risk’ we understand women who already had a first pregnancy. 85 experienced a second pregnancy resulting in a birth, 20% had the second pregnancy resulting in an abortion and 1% did not experience a second pregnancy. After 1990, 61% of women at risk had their second pregnancy resulting in a birth, 37% of women had the second pregnancy ending with an abortion and 2% did not experience a second pregnancy.

The comparison of the rates of transition to a second birth and abortion for the three periods of time (Graphic 2.9 and 2.10), shows that rate of abortion is the highest after 1990 for all ages and the lowest during 1966-1984. The rate of second birth is the lowest after 1990, and has its maximum in the early 20s (20-21) as opposed to the other two rates which have their maximum later (23-24).

Graphic 2.9. Rate of transition to a second pregnancy ending with a birth

0.300

0.250

0.200 1966-1984 0.150 1985-1989 1990-1999 0.100

0.050

0.000 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

86

Graphic 2.10. Rate of transition to a second pregnancy, ending with an abortion

0.160

0.140

0.120

0.100 1966-1984 0.080 1985-1989 0.060 1990-1999

0.040

0.020

0.000 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

87

2.2.4. Postponement of fertility

There are two main types of demographic measures: cohort and period measures.

Period measures are those taking in account the evolution of a certain demographic phenomenon during a certain period of time (usually, a year), cohort measures record the evolution of a demographic phenomenon for the entire life of a cohort. In fertility, period measures, such as total fertility rates, are more often used than cohort measures because the last ones require the cohort childbearing experience to be complete (Ni Bhrolchain,

1992:600).Period fertility is often used as estimation for cohort fertility, but it can be distorted by the changes in timing of cohort fertility. Hajnal (1947) was the first who argued that reproduction rates present a ‘misleading picture of fertility trends’ (Ni

Bhrolchain:601).Ryder was one of most important advocates of using a cohort perspective in understanding fertility changes. He also developed a formula for deducting the cohort fertility rate from the total fertility rate:

CFR=TFR/(1-c)

With c being ‘the annual increment in the cohort mean age at childbearing, with that cohort mean age assumed to be changing linearly by that amount (Kim and Schoen :4).

In an article published in 1998, Bongaarts and Feeney (1998) tried to address the problem of tempo distortion of TFR by proposing a new equation for adjusting the total fertility rate (Bongaarts&Feeney:278):

TFR’= Σ TFR’i , where TFR’i = TFRi /(1-ri)

TFRi = the observed total fertility rate at order i in any given year

88

ri = the change in the mean age at childbearing at order i.

Although their formula is quite similar to Ryder’s formula, what is new is the meaning given to the adjusted TFR (TFR’): a period, not a cohort measure that ‘corrects for distortions caused by year-to-year tempo changes’ (Bongaarts & Feeney, 2000:560).

The TFR adjusted formula was introduced in various studies (Philipov &Kohler, 1999;

Philipov, 2001) and used as an argument that the present fertility levels in several

European countries are actually higher than they seem to be.

For Romania, I computed the adjusted TFR (TFR’) using Bongaarts-Feeney formula for two years – 2000 and 2001, and it resulted that:

TFR’(2000) =1.6 , TFR(2000)=1.3

TFR’(2001)=1.5, TFR (2001)=1.2.

The difference between the observed TFR and the adjusted TFR is remarkable, with the adjusted TFR being 0.3 points higher than the observed one. Does this mean that the observed fertility is misleading, and the cohorts that are now at childbearing ages are actually postponing their children? Bongaarts-Feeney say that their formula does not try to establish a relationship between cohort and period fertility, but only to adjust the TFR for tempo distortion (Bongaarts &Feeney, 2000). In other words, there is no reason to believe that, even if we consider the adjusted TFR as being realistic, the fertility of cohorts will be higher than could be infered from the observed TFR.

In an attempt to see what the future fertility would be, I used the Reproductive

Health Survey questions regarding fertility intentions. In societies where contraceptive methods are available, women can achieve – theoretically - the number of children they want to have. It was observed that there is a strong relationship between ‘average

89

intentions’ and ‘average fertility’ (Westoff, 1990) even if at the individual level the

relationship is weak. However, as studies show (Westoff & Ryder, 1977) women in

modern societies actually have fewer children than they wanted to have.

From the Reproductive Health Survey data, it results that 55% of women without

children want to have one sometime in the future and 36.2% of those who have a child

want to have another one (see Table 2.18). These women are those who postpone having

a child: they want to have a/another child, but they will not necessarily have one in the

near future (during next year, for example). If we consider a cohort of women who have

as parity birth probabilities21 those from Table 18, then their final cohort fertility rate will

be 1.005.

Table 2.18. Percentage of women wanting another child

How many % who want children she has a/another child 0 56% 1 36.2% 26% 36.2% 43.6% 55.8% 64.4%

These data suggest that postponement does not play an important role in

Romania, and that the number of children would be lower than it is, if current desired

levels of fertility were achieved.

21 without taking in account the age: so, for example, from all women who have 1 child, 55% will have another one, no matter at what age 90

2.4.Conclusions

The following conclusions can be drawn from this analysis:

1. 1966 law had a remarkable effect on period measures in the short term and a significant influence on cohort fertility.

2. The main difference between the period before and after 1990 is that after 1990 women prefer to use abortion even for the first pregnancy, while before 1990 most first pregnancies finished with a live birth .

3. The difference in fertility rates for women with different education levels decreased after 1990.

4. Fertility after 1990 was strongly influenced by changes in population composition

(urban/rural, level of education, ethnic composition).

5. Changes in the ethnic composition of the population – a decreasing percentage of

Hungarian and Germans (which have very low fertility) and an increasing percentage of

Gypsy ( with higher fertility), due to migration and different age structures, can result in a slight increase in fertility during the next years (if there are no significant changes in the ethnic specific fertility rates)

6. Although there is a certain amount of postponement in childbearing, its role is minor in influencing the level of fertility, and it is unrealistic to believe that, after a while, fertility will increase as a result of the women catching up with the number of children they want to have

91

Chapter 3. Influences of migration on fertility

3.0. Introduction

In 1963, in his article “The Theory of Change and Response in Modern Demographic

History”, Davis argued that migration and decreasing fertility are responses that a

population has in times of extraordinary hardship. Although his article refers to high

fertility settings, his ideas make sense for other situations too, generally speaking, for any

population that experiences a high level of poverty. The experience of Eastern European

countries after 1990 seems to be a good example for Davis’ theory: living in declining

economies (Table 1), during times of extreme economic uncertainty, people migrate and

reduce their number of children.

Table 3.1. GDP in Eastern European countries, 1989-1992 (1990 Geary-Khamis Dollars)

Bulgaria Czechoslovakia Hungary Poland Romania USSR Yugoslavia 1988 6336 8675 6929 5790 4035 7032 6026 1989 6217 8729 6787 5685 3890 7078 5917 1990 5764 8464 6348 5113 3460 6871 5458 1991 4516 7244 6010 4798 2887 5793 4861 1992 4054 6845 5638 4726 2565 4671 3887

source: Maddison, 1995

Although high migration and declining fertility can be seen as independent outcomes of the same cause (high poverty) there are certain relationships between the two outcomes. Several studies (Lindstrom& Saucedo, 2002, Singley & Landale, 1998,

Hervitz, 1985) investigated the relationships between fertility and migration for the case

92 of people migrating from countries with high fertility to countries with lower fertility in various regions of the world. Some researchers suggest that migration is a disruptive process for fertility: migrants tend to postpone having children because of the socio- psychological stress associated with living in a new place and spouse separation. On the other hand, as other studies pointed out, migrants are not a random sample from the origin country population, they are selected, they have special characteristics and having fewer children is one of their features. Migrants, said other students, tend to adopt – gradually or quickly- the norms of the receiving country in many aspects, including number of children, and after a while their fertility resembles more the fertility of the destination country rather than that of the sending country.

This chapter analyzes the relationship between international migration and fertility, after 1990, in Romania. Using various sources of data (Census data and surveys), I am trying to answer to the following questions: 1) are migration and low fertility interdependent? and 2) does migration distort fertility measures?

93

3.1. Migration and fertility. Interrelationships

Migration is a spatial phenomenon (Hammar&Tomas: 15), migrants being people who change their residence for a period of time. Even if “migrant” and “migration” seem to be easily understandable, the meaning can vary from study to study. Usually, refugees

(people who were forced to migrate) are considered a distinct and special group of migrants, but even within the non-refugee migrants, the definitions can be very different.

For example, under US law, an immigrant is a foreigner who is entailed to live and work permanently in the United States (Martin&Midgley: 5), but many studies discuss

“undocumented migrants to the US” (Massey&Espinosa: 940), obviously not taking into account the US legal definition of immigrant.

Some researchers also consider the distinction between international and internal migration to be important, while others think it is rather a formal one (Hammar&Tomas:

15). International migrants are defined as people who change their residence from one country to another (they crossed at least one national border); internal migration is defined as migration within national borders.

Several micro and macro level theories try to explain how internal and international migration began and why they continue in the world. Taking into account to what moment of migration they refer to, these theories are often classified in two main groups: initiation and perpetuation theories.

Theories of the initiation of migration fall into four main perspectives: neoclassical economics, new economics of migration, dual labor market and world system theory. The second group (perpetuation) includes network theory, institutional theory, migration

94

systems and cumulative causation theory. Neoclassical economics theory, from a macro

level point of view, pointed out the geographical differences in labor supply and demand

as being the main cause of migration (Massey& all: 433). At a micro level, the neoclassical economic theory is a model of individual choice based on a cost – benefit analysis (Massey&all:434).Potential migrants estimate the cost and future benefits they would have from moving to another place, and choose the place where they can get the best returns (Massey&all:434)

The new economics of migration theory does not consider the migrant as an isolated individual, but as a part of a household (family). The migration decisions are made within the household and the people involved in them try to minimize different risks associated to the household (crop market fluctuations, unemployment) or to enable families to capital necessary for financing new projects (Massey& all: 436).

Dual labor market theory (or segmented labor market), developed by Piore, argues that immigration is not a result of the sending country conditions, but an effect of the industrial society’s characteristics. Wages are not simply the result of supply and demand, they are also embedded in a hierarchy, and they are attached to a social status. In a dynamic economy, there will always be a need for a work force at the bottom of the hierarchy, because people who used to be there advance on the social scale. The need for low skilled work cannot be simply covered by increasing the wages attached to this work, because this will upset people who are doing more skilled work and get a lower wage.

Two categories of native population are used to fill the need of low skilled work: women and teenagers (Massey & all: 423). In a less advanced society, women work mostly part – time, and their social status is not defined by the work that they do, but by the position of

95

their husband. The situation changed in nowadays-industrialized societies, where women

have more stable jobs, and are more interested in their careers. Teenagers also used to be

a segment of population that accepted low skilled jobs, but because of the drop in birth

rates and increase in formal education they are not enough teenagers anymore to fill all

these jobs. In this way, the immigration remains the only solution for filling the low paid

jobs in an advanced economy. The immigrants do not consider themselves as a part of the

host, but of the sending society, so they do not feel ashamed to be at the bottom of the

host country hierarchy. They accept jobs and wages that the native would not accepted,

helping in this way the receiving country economy to go up.

World system theory considers migration to be a result of the world market

structure (Massey& all: 444). The incorporation of more and more land and population

into the world market economy because of capital penetration into peripheral societies

and global inequality as a structural given (Zolberg:403) makes more people migrate

from a region to another.

One of the most popular theories trying to explain the perpetuation of migration

nowadays is network theory. Migrants, former migrants and non-migrants in origin and host countries are related through a set of interpersonal ties that facilitate migration by reducing the costs of moving and risks. As a result, once a migration process began, the migration stream between two countries tends to expand and is less likely to be controlled by the governments, because the interpersonal networks are outside of their possible control (Massey& all: 450).

Institutional theory predicts that, once a migration process begins, private institutions that offer different kinds of services for migrants arise. Clandestine transportation,

96

counterfeit documents and visa, lodging in the country of destination, and many other

services make migration and the adaptation of migrants in the receiving country easier

(Massey& all: 450).

Cumulative causation theory argues that each act of migration changes the social

context, making further movements more likely to occur (Massey& all: 451). Migration

affects the distribution of land, income, and human capital, the organization of

agriculture, culture and the social meaning of work. The relative deprivation hypothesis

(people who feel poorer in comparison with their neighbors are more likely to migrate) is an example of how migration affects the distribution of income and is further affected by the changes induced in the distribution of income (Massey& all: 452).

Migration interferes with other demographic processes, such as mortality and fertility.

Immigrants are selected populations coming from countries with a different culture than the receiving country has. Their mortality patterns can be very different than those of the native people, as the case of Mexicans in the US shows: Mexican groups in the US have one of the lowest infant mortality rates, and mortality at old ages is lower than the mortality of non-Hispanic White population, despite their low socio-economic status.

Immigration, on the other hand, is considered a possible solution for the problem of aging in the developed world (Jonsson & Rendall: 129). Although in the long term it is assumed that immigrants fertility will not differ from the native population fertility

(Espenshade, 1994) so immigration does not have a direct effect on fertility, in the short term immigration brings an infusion of young population and delays the aging of a population.

97

Four main hypotheses regarding the relationship between migration and fertility

have been developed. The selection hypothesis considers that migrants are a special

group, with characteristics that differentiate them from the non migrants. Having fewer

children – or norms regarding a low fertility – is one of these characteristics that

differentiate them from the non migrants (Singley&Landale:1439). The disruption

hypothesis suggests that migration lowers the fertility of migrants because of spousal

separation and because moving to a new place is a stressful event, making migrants

postpone having children until they adapt to the new country. The assimilation and

adaptation hypotheses predict that the fertility of migrants will become – sooner or later

– similar to the fertility of the receiving country population. These two last hypotheses were tested only for the case of migration from countries with high fertility to countries with low fertility. It would be interesting to see if migrants moving from low fertility countries to countries with higher fertility (as Eastern Europeans to Western Europe) tend to have higher fertility than the non migrants from the origin countries. Most of the studies investigate the relationship between fertility and migration in receiving countries - studies on the influence of emigration on the fertility of the sending countries are very rare, because in most cases sending countries have a high fertility and the levels of out- migration have little if any effect. Ebanks etal. (1975) investigated the influence of emigration on fertility in Barbados, and they concluded that emigration played an important role in the fertility decline of the country. Lindstrom& Saucedo (2002) studied the short and long term effects of US migration experience on Mexican women’s fertility in Mexico and observed that spousal separation reduces the probability of a birth in short term and the fertility of Mexican women migrants in the US is lower because of

98

migration. The case of Eastern European countries, with below replacement fertility and

countries of emigration is quite singular. Philipov (2001) was among the first who

advanced the hypothesis that high emigration of Eastern Europe (especially South-

Eastern Europe) leads to underestimates of fertility, because the official statistics are

calculated without taking emigration22 into account and, in this way, fertility is

underestimated. He argued that, in the case of Bulgaria, where 8% of the population is

estimated to work and live outside of the country (Philipov:9), the TFR in 1998 would be

0.2 higher if migration were taken in account. Migration could also have an impact on

family formation, as Sklar (1974) argued for the case of Central Europe, where the

massive emigration of young men in the early 1900s made ‘increasingly difficult for

women in the Czech, Baltic and Polish areas to find husbands’ (Sklar:241) and, as a

consequence, age at marriage and number of single women increased.

22 Official statistics take into account only those emigrants who establish officially their residence outside of a country. However, most of the emigrants from South Eastern Europe are circulatory migrants, who work outside of their native country in many cases illegally, so their number cannot be computed from official statistics, but only estimated. 99

3.2. East –West Migration in Europe – levels and theories

a) East-West migration before 1990

The economic East –West migration in Europe was a negligible phenomenon during

1945-1989, because most of the communist governments did not allow migration between their countries and Western European countries. However, before 1950, the phenomenon had quite a history. During 1850-1920, when almost 30 million people left

Europe for North or South America, several hundred thousand Polish and Ukrainians migrated to France, Germany or England (Fassman& Munz, 1994:520). About 3% of the

Czech population and 6% of the Polish population migrated during 1900-1914

(Sklar:241). Three factors are considered important in explaining this trend: the economic growth of the West, due to the industrial revolution, differences in political systems between the East and West, and the rise of nationalism in Central and Eastern Europe, which forced ethnic and religious minorities to migrate (Fassman&Munz, 1994: 522).

During 1944-1989, migration from Eastern Europe had two main components: migration of people subjects to special political agreements (migration of ethnic minorities) and economic migration. In forty years, between 12 and 14 million people migrated from Eastern to Western Europe, most of them emigrated because of their ethnic origin.

Before 1918, some regions of Eastern Europe were part of the Habsburg Empire, and many Germans settled down in these regions. After 1918, when Habsburg Empire was divided into independent countries, some of these ethnic Germans remained in the newly

100

formed countries. Immediately after 1944, when Germany was defeated, some of the

Germans living in Eastern Europe were dislocated and sent to Germany23. That was a

first wave of Germans leaving Eastern Europe. Later on, Federal Republic of Germany,

negotiated bilateral agreements with some of the Eastern European countries (including

Germany Democrat Republic), allowing ethnic Germans living in those countries to

move to if they wanted to. In return, the Eastern European countries received some

financial support from FRG (Federal Republic of Germany). Similar agreements were

negotiated by Israel, and Turkey. As a result, during 1960-1992, about 500,000 people

(mostly Jewish) moved from Eastern Europe to Israel or United States. During 1946-

1955, over 200,000 people left Bulgaria (Guentcheva&all:20), many of them members of

ethnic minorities (Turks, Armenians, Jews, Czechs and Slovaks) and settled in their

country of origin or in other countries.

Although there were mainly economic reasons behind these people emigration, they

were able to leave only because of these special political treatises, so they are not

considered either economic migrants or refugees. Among the former communist

countries, only Yugoslavia allowed its citizens to work and immigrate to a capitalist

country before 1989, and this is why only 15% of East-West migrants during 1950-1989

can be classified as labor migrants or as dependent family of labor migrants

(Fassmann&Munz, 1994: 527).

23 This was the case with the Germans living in Czechoslovakia, where the population suffered intensively during the Second World War because of the German occupation. The German minority from Czechoslovakia helped the German enemy, so when the War ended, the native population was very angry at their German fellows. As a result, most Germans were deported to Germany, and their possessions confiscated by the Czechoslovakian government. 101

b) East West migration after 1990

After 1990, three main groups of East-West migrants can be identified: ethnic

Germans moving to Germany, refugees from former Yugoslavia and economic migrants.

The number of ethnic Germans (Aussiedler) moving from Eastern Europe to

Germany increased, because the new Eastern European governments made easier the

emigration of this category (Table 3.2).

Table 3.2. Number of ethnic Germans living in Eastern Europe who migrated to Germany after 1990

Year Country of origin 1990 1995 1996 1997 1998 1999 2000 2001 2002 Total 397,073 217,898 177,751 134,419 103,080 104,916 95,615 98,484 91,416 Poland 133,872 1,677 1,175 687 488 428 484 623 553 Hungary 1,336 43 14 18 4 4 2 2 3 Romania 111,150 6,519 4,284 1,777 1,005 855 547 380 256 Former Czecheslovakia 1,70862148 1611182213 Former Soviet Union 147,950 209,409 172,181 131,895 101,550 103,599 94,558 97,434 90,587 Former Yugoslavia 961 178 77 34 14 19 -- 17 4 Other countries 96 10 6 -- 3 -- 6 6 --

source: Statistisches Bundesamt (Federal Statistical Office), apud Migration Source

The disruptive Yugoslavian war after 1990 made many Serbs, Croats and Bosnians migrate and ask for refugee status in Western Europe (Table 3.3). Because Yugoslavia allowed their citizens to work in Western countries before 1990, many Yugoslavian had ties in Western Europe, and that made easier for them to emigrate and ask for refugee status when the war began. After the war (or wars) was over, the economic situation of some of the former Yugoslavian republics (Bosnia Herzegovina, or even Serbia) prevent many of them from coming back to their countries.

102

Table 3.3. Stock of immigrants from former Yugoslavian republics in European countries Population of foreign citizenship in Country of Germany Austria Italy Belgium citizenship (1.1.1999) (1.1.1999) (1.1.1998) (1.1.1999 Bosnia and Herzegovina 281,380 10,246 1,995 Croatia 206,554 13,575 736 Former Yugoslavia* 721,029 337,863 74,869 12,130

source: Recent demographic developments in Europe, 1999

* Serbia, Montenegro, Croatia, Slovenia, Bosnia-Herzegovina

In the group of economic migrants, Romanians, Bulgarians, Polish and Albanians are well represented24 left their countries and went to Western countries after 1990.

Immediately after 1990, many of them applied for a refugee status, as the old laws (which

allowed citizens of Communist countries to ask for refugee status) were still enforce. The

difference in wages between Eastern and Western Europe is an important reason of these

migrants, so the neoclassical theory of migration can offer a good explanation for why

these people moved. However, these emigrants did not go to the countries with the

highest wages, as it was expected, but preferred countries with rather low wages, as

Spain, Portugal and Italy.

. The pattern of migration (which countries send people in what countries) it is an

important point in understanding East-West migration in Europe. The direction of

migration (where people from a certain country migrate) can be explained from several

points of view. First, geographic vicinity plays an important role: people tend to move to

24 Russian immigration was also feared by Westerners: in a study done in 1993, demographers estimated that, in the future, 1.5 million Russians plan to emigrate to Western Europe (Shevtsova: 244), and this seems to be an underestimation given the fact that, in 1991, 50% of Russians declared they were ready to leave the country (Shevstova: 244) 103

states that neighbor their native country. This is the case with some waves of migration in

Eastern Europe (Romanians of Hungarian origin into Hungary, Bulgarians with Turkish

origin going to Turkey, Albanians going to Greece, Polish going to Germany), although it

does not work for other emigrant groups. A second hypothesis argues that cultural,

political and historical connections between the origin and host country determine the

direction of migration (Fassman & Rainer, 1992:469). This hypothesis can explain the

emigration of Romanians in Italy, Spain and Portugal, countries that share with

Romanian a Latin background in languages and cultures.

Figure 3.1. East-West main directions of migration in Europe, after 1990

One group of migrants that received special attention during the last years are Gypsy, maybe because they are the only one groups with special racial and cultural traits that can be easily identified. Between January and September 1991, a first wave of about 70,000

104

Gypsies from Romania and another couple of thousand from Macedonia reached

Germany (Barany: 243), and they were received with high hostility by the local

population. Later on, groups of Gypsies went to Finland (many of them from Bulgaria),

England (from Slovakia), Ireland (from Romania), France, Italy and Spain (Romania and

Bulgaria) and Austria (Slovakia). Although they are not very numerous – far less

numerous than the African migrants – mass media and population reacted very strongly

to their presence on the streets and town of Western Europe25. Eastern European Gypsies

are even rejected by the Western Gypsies, maybe because they are much poorer than the

last ones.

As an overall, the number of migrants from Eastern European countries was much

lower than it was feared, (Lazaroiu, 2003, Guentcheva, 2003, Korys, 2003). The benefits

that Western Europe has by getting foreign people to work for low wages seem to surpass

the negative effects of having immigrants, and this is why the immigration regulations

regarding Eastern Europeans became more and more relaxed. On the other hand, the

remittances sent by immigrants back to their countries of origin play an important role in

the GDP of these countries. For Bulgaria it was estimated that in 2002, emigrants sent

home 449.6 million USD, which represent 2.9% of the Bulgarian GDP, and surpass the

amount of direct foreign investment by 20.9 million (Guentcheva&all:16). National Bank

of Romania estimates that, during 2002, the emigrants transferred home 1.2 billion USD

(Lazaroiu:20). Poland, with an estimated number of 786,100 emigrants, benefited from

900 million USD in remittances in 1998 (Korys: 40).

25 French newspapers used to have large articles about Gypsies in France, and their involvement in prostitution and human traffic. 105

However, the citizens of the former communist countries who joined EU in 2004 still face restrictions in working in some of the old EU countries. These restrictions do not seem to be the result of a careful analysis regarding migration, but rather the result of the

Westerners fear of the EU enlargement effects. Some of the Eastern European countries – as Czech Republic and Slovenia – actually became countries of immigration after 1995, and their citizens have hardly any reasons to move massively toward West, when the economic growth moves East in Europe.

106

3.3. Data

While emigration has positive effects on the economic growth of some of the Eastern

European countries, there are also negative effects to be analyzed. Romania is one of the countries with high migration and below replacement fertility. UN studies placed

Romania among the first ten countries of emigration with an estimated 80,000 emigrants per year (around 1995). At the same time, fertility attained 1.3 children per woman in

2003, a level well below replacement. The last census (2002) shows that between 1992 and 2002, the Romanian population decreased by one million (~5%) as a result of migration and negative natural increase. It can argued that the high emigration rate has a negative, disruptive effect on fertility, because of spousal separation and because moving to a new place is a stressful event, making migrants to postpone having children until they adapt to the new country.

There are two types of data available for studying the relationship between migration and fertility: censuses and surveys. Migration became significant only after 1990, so I will use 1992 and 2002 censuses to estimate the number of migrants and the influence of migration on fertility. For survey data on migration, I will use “Community census on temporary external migration of rural population”, a survey conducted by International

Migration Organization, University of Bucharest, Ministry of Public Information and

Ministry of Internal Affairs in 2001. There were 12357 villages (from 127000) and 148 small towns (from 152) included in this study. For each village and small town information was recorded on the number of emigrants, on people who worked abroad and

107 came back, and some characteristics of the community (Sandu, 2001). People involved in the study recognized that data can be affected by errors, taking into account that the study was done on the village/small town level not on the individual/household level (Sandu,

2001), and they suggested that data at the county/region level are much more reliable than data on the village/town level. This study does not have any data about fertility, so for studying the relationship between migration-fertility, I used fertility data from annual publications of Romanian National Institute for Statistics, Romanian Yearbook (2001) and National Census (1992).

108

3.3.1. Census data

One method for estimating the number of out-migrants from a country with low

immigration is by using data from two consecutive censuses. For the relationship

between fertility and migration, it is particularly important to estimate the number of

women migrants who are in the reproductive ages.

By comparing the female population ages 15-64 in 1992, on age groups and

nationalities in Romania26, with the female population age 25-74 in 2002, I find that for

all ethnic groups except Gypsies, the female population decreased (Table 3.4).

Table 3.4. Difference in size between 1992 and 2002 (female cohorts age 15-64 in 1992)

Total Romanians Hungarians Gypsy Ukrainians Germans 15-19 101224 88033 10805 -480 618 2156 20-24 56147 46471 9077 -1547 304 2023 25-29 30828 26828 4164 -1081 149 1141 30-34 42151 37353 4377 -987 144 1411 35-39 44282 38796 4431 -684 149 1577 40-44 33697 29300 3573 -354 144 1039 45-49 39709 32670 4891 335 206 1475 50-54 71367 61296 6084 800 274 2324 55-59 100221 85771 8727 1602 336 2751 60-64 138001 118959 12328 1579 462 3082

The number of Gypsy females increased for all age groups between 1992 and

2002, which is not an expected result there is no evidence of Gypsy immigration to

26 In this chapter I focus only on out migration. In 2002, there were 23,945 foreign citizens living in Romania for 12 months and over (Romanian Census, 2002).There were also 109,924 Romanian citizens born in other countries (Romanian Census, 2002). Most of these people lived all their life in the same place, but because of the frequently changing borders during the 20th century they are recorded as being born in another country. In the following I will assume no immigration. 109

Romania. Actually countries as France, Italy and Spain claim to have an increasing stock of Romanian Gypsies.

There are two possible explanations for this result. Figures regarding the number of Gypsies are not very reliable anywhere and there are still cases of children born at home who do not have birth certificates. It is also very possible that the counting was not accurate in 1992, or in 2002 or in both years. Another reason can be related to how ethnicity is recorded: the respondent identifies himself/herself as belonging to a certain ethnicity, so people can declare themselves as being Gypsies at one point in time, and then declare some other ethnicity. Usually, Gypsies choose to declare themselves as

Romanians, but there are cases when they declared other ethnicities. One well publicized case of such shifting in identity appeared in Transylvania when the Hungarian government decided to give some special rights to Hungarian ethnics living abroad (so called Status Law). The so called Hungarian Gypsies – a Gypsy tribe claiming to be similar to Gypsies living in Hungary - living mostly in the Western part of the country, declared themselves as Hungarians in order to qualify for the Status Law rights, and got very angry when the Hungarian associations denied they were Hungarians. During the past years, several policies aimed to make Gypsies more aware of their own ethnic identity and to take away the blame from being a Gypsy. Probably the results of the 2002 census show that these policies were successful. The data problems with counting

Gypsies made me use, in the following, only the other four ethnic groups (Romanian,

Hungarian, Ukrainian, German).

In order to estimate the number of women migrants, I computed (from 1992 figures) the number of women there would be in 2002 in the absence of migration. As

110

migration occurs usually at young ages, I used only the cohorts that were 15-34 in 1992,

and I estimated the number of deaths for these cohorts using the death rates from the life

tables published by the Romanian National Institute for Statistics (Demographic

Yearbook, 2001)27. I considered all ethnic groups as having similar death rates; as the

death rates were not very high (see footnote 3), they would not be significant even if

there would be differences in the death rates. Table 3.5 shows the number of women that

should be alive in 2002 in the absence of migration, and the number of women that were

alive in 2002, for the four ethnic groups.

Table 3.5. Number of women that are that should be alive in 2002, in the absence of

migration

Romanians Hungarians Ukrainians Germans without without without without Age in 1992 actual actual actual actual migration migration migration migration 15-19 817570 744186 62159 52468 2793 2225 3435 1341 20-24 892063 866261 65778 58225 2251 1999 3691 1754 25-29 539122 529674 39503 36612 1924 1837 2199 1129 30-34 658464 655462 46953 45025 2008 1969 2886 1626

These data show that, among women who were age 15-34 in 1992, and belonging to

one of the four ethnic groups (Romanian, Hungarian, Ukrainian, and German), about 4%

migrated by 2002 – 5% of the Romanian group, 12% of the Hungarian groups, 2% of the

Ukrainian group and 72% of the German group emigrated. Let us consider the worst

scenario, with these migrants never having children in Romania during 1992-2002.

Suppose that these women would have the same fertility as the women who did not

27 I had the death rates only until 2000, so I assumed they remained until 2002. The death rates were between 0.002 and 0.008. 111 migrate if they stayed. Then, for each 1000 women who migrated, we can estimate the number of children lost as the difference between the numbers of children ever born in

2002 and in 1992 to women who stayed (Table 3.6).

Table 3.6. Number of children ‘lost’ because of migration, for every 1000 women,

1992-2002

age in 1992 Romanians Hungarians Ukrainians Germans 15-19 877.4 852.9 1299.3 666.3 20-24 765.7 839 1263.1 769.4 25-29 348.9 395.1 761.1 334.2 30-34 128.6 142.7 384.3 160.8

Table 3.7. Estimated number of women-migrants

Romanians Hungarians Ukrainians Germans

15-19 73384 9691 568 2094 20-24 25802 7553 252 1937 25-29 9448 2891 87 1070 30-34 3002 1928 39 1260

From Table 3.6 and 3.7, about 108, 430 children were ‘lost’ because of female migration. Taking into account that, during 1992-2002, there were around 2.4 million children born in Romania, it means that the fertility would be at most 5% higher if women did not migrate during this interval of time.

Considering the second issue, that high migration causes the fertility measures to be underestimated, I compared the TFR for 1999-2001 from official statistics with the

TFR computed with 2002 census data. Using the 2002 census data and probabilities of

112

death for women age 15-4928, I projected backwards the number of women that would be

alive in 2001, 2000 and 1999 for each year of age. Then I computed the TFR for 1999-

2001 using figures resulting from the projection as the denominator 29 and I compared

them with data published by the National Institute for Statistics. The results are presented

in Table 3.8.

Table 3.8. TFR and number of women age 15-49 years

Women, age 15-49 TFR Used in Used in Year Projected Projected statistics statistics 2001 5538580 5817197 1.32 1.2 2000 5543080 5816755 1.38 1.3 1999 5549933 5817083 1.41 1.3

As it can be observed, TFR is 0.1 higher when the backwards projected number of

women is used as the denominator because the number of women age 15-49 used in

official statistics is about 5% smaller than the number of women resulting the projection.

The difference is probably a result of not taking out migration into account the official

statistics, so, as Philipov (2001) argued for the Bulgarian case, there is certain bias in

measuring fertility in Romania, due to a high rate of out migration. As these results

showed, the TFR would be around 8% higher if migration is taken into account.

28 For probabilities of death, I used the complete life table on sex, 1998-2000, the only one with data for each year of age. 29 For number of births I used published data (Births, 1999, 2000, 2001, Romanian National Institute for Statistics). 113

3.3.1. Survey data

One of the few studies that examined the migration process using data from a country of origin was done in Romania in 2001. Using a so called “community census” –in which for each village, data were recorded regarding the characteristics of the village, the history of internal migration and international migration - the authors were able to estimate the volume as well as the main directions of economic migration from Romania to European Union.

Data were recorded about temporary migration during 1990-2001 for people from

12300 villages (from a total 12700 villages) and 148 small towns (from 152). The results showed that the large villages, with a high level of development, located especially in

Transylvania or in the Western part of the country have the highest level of international out migration.

Six countries seem to absorb most of the migrants from Romania: Yugoslavia,

Hungary, Germany, Turkey, Italy and Spain. In the late 1990s, Germany is no longer a preferred destination of Romanian migrants; Italy and Spain now capture more and more migrants with few of them returning. Yugoslavia and Hungary have a special position; people travel frequently in these countries for buying and selling things, but also for working there. Gypsies (a category frequently mentioned in the articles about migration in Western Europe), go especially to Germany, Yugoslavia and Spain.

From a demographic, as well as an economic point of view, Romania is not a homogeneous country. During 1945-1990, the communist governments tried to reduce

114 the economic disparities between regions by industrializing and urbanizing the poorest parts of the country (Moldavia and the southern part of Wallachia, Figure 3.2).

Figure 3.2. Historical regions in Romania

Historically divided in regions (Transylvania, Moldavia, Banat, Crisana-Maramures,

Oltenia, Wallachia, Dobrogea), the Romanian administrative structure changed several times during the last one hundred years. The concepts of the main administrative units

(village, comuna, town and county) remained unchanged during the last forty years, although the number and boundaries of counties and towns changed frequently. After

1990, the administrative changes were minor, with the rural region around Bucharest becoming a county and being separated from the main town.

The forty one counties that now compose Romania have different demographic and economic characteristics: those situated in the North Eastern part are among the least developed and have the highest fertility rates, while those in the Western part and the capita (Bucharest) are the most developed and have the lowest fertility rates (Table 3.9.).

115

Table 3.9. Demographic indicators in the poorest and richest counties from Romania

Rank of Life expectancy, Life expectancy, County Region development, TFR 1989 TFR 2000 males, 1988-1990 males, 1998-2000 1998* Vaslui Moldova 1 2.16 1.8 68.01 67.92 Botosani Moldova 2 2.97 1.8 65.98 66.17 Suceava Moldova 3 2.57 1.7 68.02 68.57 Sibiu Transylvania 37 2.02 1.2 67.24 67.14 Cluj Transylvania 38 1.91 1.3 67.01 68.22 Timis Banat 39 1.86 1.2 65.57 66.93

Brasov Transylvania 40 1.87 1.1 67.34 67.75

source: Demographic Yearbook and The villages of Romania, World Bank Report

In terms of out-migration, Romania is not homogeneous either: 4.5% of all villages

contain 60% of all returned migrants and 20% of all migrants (Sandu:26), which shows

the emigration phenomenon is regionally concentrated. Most migrants come from the

North Eastern and Western part of the country. There are three major groups of migrants:

one group is the migrants leaving the Central-Western counties (Covasna, Harghita,

Mures, Salaj), many of them with Hungarian origin, who go to work in Hungary because

they know the language and also have various benefits30. As statistics show, the most

numerous group of foreign citizens living in Hungary are those coming from Romania:

62,130 on January 1st, 1998, which represent about 42% of all foreign citizens living in

Hungary (Recent demographic developments in Europe, 1999). A second group is formed

by those coming from the Northeastern counties (Moldavia), who go for work to

30 Status Law gives various benefits to ethnic Hungarians living outside Hungary who come and work in Hungary: they get an work permit very easily, and the state pays the health insurance. A person who works continuously for three years in Hungary is eligible to apply for permanent residence. 116

Portugal, Italy and Spain. A third group of counties (Timis, Arad, Sibiu, Brasov) send

most of the emigrants to Germany. Gypsies from Transylvania migrate to Germany (26%

of people who came back from Germany are Gypsy), Hungary (15% of all people who

came back from Hungary are Gypsy) and Spain (Sandu: 25). Although French media

concentrates on Gypsies’ migration in France (from Romania as well as other Eastern

European countries), in the general picture this group is not numerically important: only

4% of people who came back from France are Gypsy.

The relationship between migration and fertility in Romania can be studied from

two points of view: 1) emigration has a disruptive effect on fertility (and, in this case,

those counties with the highest levels of migration would also be those with the lowest

levels of fertility) and 2) emigrants postpone having children until they get back from

working abroad (and in this case, the counties with high rates of returning migration

would have higher fertility). For these hypotheses there are several arguments: migrants

are, generally speaking, young men and women who, by going to work abroad, postpone

marrying and having children. On the other hand, it can be argued that people postpone

having children and migrate because they live in poverty and want to get a better

economic situation.

In the following, I tested these hypotheses at the county level. For migration, I

used two types of measures: a ‘rate of out migration’: the number of people from a

county who, at the time of the interview, were working abroad divided by the population

of that county (measured at 1st of July 199731) and a ‘rate of return’: the number of people

31The ethnic distribution of population in Romania is available only for the census years - 1992 and 2002. For this study, because data regarding migration refer to whole period between 1990-2001, for the proportion of Gypsies I used an average of the census figures for 1992 and 2002. Theoretically, this would 117

who came back in the county after working abroad during 1990-2000 divided by the

population of county (measured at 1st of July 1997).

For fertility, I used two measures: Total Fertility Rate in 2000 (TFR2000) and the

difference between Total Fertility Rate in 1989 and Total Fertility Rate in 2000 (TFRD).

Among the three most numerous ethnic groups (Romanians, Hungarians, Gypsies),

Gypsies tend to have higher fertility, while Romanians and Hungarians have similar

levels (Hungarians tend to have fewer children, but the difference between Romanians

and Hungarians is not significant as data from censuses and other studies show) (see

Table 3.10).

Table 3.10. Fertility of Romanians, Hungarians and Gypsies in Romania

TFR* CFR1 CFR2

Romanians 1.2 2.07 1.84

Hungarians 1.3 2.03 1.82

Gypsies 2.6 4.03 3.71

source: Reproductive Health Survey, 1999 and Romanian Census, 2002

*- TFR was estimated for 1994-1996 from survey data

CFR1 – cohort fertility rate, measured in 2002, for women age 40-44 in 2002

CFR2 – cohort fertility rate, measured in 2002, for women age 35-39 in 2002.

Although it is not the final cohort fertility, because women will still have children, I thought it could be useful to show that the differences in fertility between ethnic groups continue to manifest even for young women.

be the proportion of Gypsies in the middle of the interval 1992-2002, which is July 1, 1997. This is why I used population at July 1, 1997 for the other variables included in the analysis.

118

Additionally I controlled for several variables:

- ethnic composition of the county measured with the variable:

o Gypsy (proportion of Gypsy in the county in 1997- see note 7)

- economic performance of the county: - rate of unemployment in 1997

- location of the county (Moldavia is historically a region with high rates of fertility)

- age at first marriage for women (age at marriage is considered to play an important role in fertility)

- fertility in 1989 (I used this factor to include the unmeasured things that remained constant after 1989: values related to fertility, ideal number of children, etc)

The data regarding migration have two components: 1) people who left after 1990 and are, at the moment of the survey, working abroad, and 2) people who worked abroad for a while after 1990, but came back later on, and are living in Romania at the moment of the survey. In order to test the influence of migration on fertility, I used two models. In both these models, I included in the analysis all Romanian counties, so this analysis in done at the population, no sample level. This is why there is no need for testing if the coefficients are significant or not.

a) Relationship between fertility and out migration

In the first model, I tried to estimate the relationship between out migration during

1990-2000 and fertility change during this period of time. I expected to find a negative relationship between the fertility change and out migration (the higher the rate of out

119 migration, the higher the decline in fertility). This result would reinforce the idea of migration as having a disruptive effect on fertility. I measured fertility change with the difference between the total fertility in 2000 and the total fertility rate in 1989 as the dependent variable. Fertility declined for all counties during 1989-2000, so TFR2000-

TFR1989 is a negative variable (varying between -0.36 to -1.3). I measured migration with a variable called ‘rate of migration’: total number of people who worked abroad during the interval 1990-2001, divided by the population at July 1, 1997. I controlled for fertility in 1989 as a measure for all ideas regarding children and family that eventually remained unchanged after 1990.

The regression coefficients for this model are reported in Table 3.11.

Table 3.11. Regression coefficients

Unstandardized Standardized Coefficients Coefficients B Std. Error Beta (Constant) 0.449 0.181 %Gypsy 0.020 0.015 0.147 Moldova 0.185 0.059 0.404 Unemployment 0.003 0.007 0.045 Change in age at marriage -0.053 0.056 -0.112 Rate of migration -0.001 0.001 -0.079 TFR89 -0.602 0.068 -0.920

R2 adjusted=0.66,

N=41

TFR1989 : total fertility rate in 1989

%Gypsy: percentage of Gypsy population in the county (1997, see note 7)

120

Moldavia: dummy variable, with 1 if the county is located in Moldavia and 0

otherwise.

Unemployment: rate of unemployment on 1 July 1997

Change in age at marriage: the difference between age at first marriage for women

in 2000 and age at first marriage for women in 1989.

Percentage of Gypsy in a county has a positive influence on the difference in

fertility: 1% increase in the percentage of Gypsies in a county increases the change in

TFR during 1989-2000 with 0.147. However, because the dependent variable has only

negative values, an increase in the dependent variable represents a decrease in the

absolute value of the dependent variable (for example, if the dependent variable increases

with 0.147 from -1.3, it means it will get the value of -1.153, and -1.3< -1.153, but -

1.3>-1.153. In other words, a higher percentage of Gypsies in a county makes the absolute change in fertility during 1989-2000 to be smaller.

A higher rate of migration is associated with a decrease in the dependent variable, but, in absolute terms, it makes the gap between fertility in 2000 and in 1989 to be wider.

So out migration has a disruptive effect on fertility, as those counties with higher rates of out migration are also those who recorded the largest decrease in fertility.

b) Relationship between fertility and return migration

In the second model, I wanted to see what is the relationship between returning

migration and fertility. My hypothesis is that a higher density of returning migrants

makes the fertility to increase, because these migrants have now the financial means to

121

establish families, so they have no reasons to delay having them. The dependent variable

in this model is total fertility rate in 2000 (TFR2000) and the independent variable is:

‘rate of return’ (number of people who worked abroad and came back, divided by the

total population (measured at July 1st, 1997)). The coefficients are presented in Table

3.12.

Table 3.12. Regression coefficients for analysis, fertility

Unstandardized Standardized Coefficients Coefficients B Std. Error Beta (Constant) 1.525 0.973 TFR 1989 0.364 0.074 0.542 %Gypsy 0.022 0.015 0.153 Moldova 0.210 0.056 0.448 Rate of returning migration 0.001 0.004 0.024 Rate of out migration -0.003 0.003 -0.117 Unemployment 0.002 0.007 0.024 Age at first marriage, 2000 -0.046 0.038 -0.141

R2 adjusted=0.67,

N=41

I controlled for additional variables:

Rate of out – migration: number of people who were working abroad at the

moment of the survey divided by the population32

TFR1989 : total fertility rate in 1989

%Gypsy: percentage of Gypsy population in the county. I used an average of 2002

and 1992 census data for this variable- ethnic composition of population is available only

32 I used population at 1st of July, 1997, from Romanian Yearbook, 1997, published by Romanian National Institute for Statistics. 122

for census years (which, theoretically, would be proportion of Gypsy population in the

county in 1997).

Moldavia: it is a dummy variable, with 1 if the county is located in Moldavia and 0

otherwise. Moldavia is historically a region with high fertility rate

Unemployment: rate of unemployment on 1 July 1997

Age at first marriage, 2000: age at first marriage for women in 2000.

As can be observed from the standardized coefficients, the total fertility rate in

1989 and being locate in Moldavia play the most important role in the fertility level in

2000. An increase in the age at first marriage is associated with a decrease in fertility; this

is not an unexpected result, as many fertility studies show that fertility and age at

marriage are negatively correlated.

The rate of return migration, controlling for all other factors, has a positive influence

on fertility, while the rate of out-migration decreases fertility. Migration has a disruptive

effect on fertility: young people who leave the country to go to work abroad postpone

marrying and having children until they will have a better financial situation33; the negative coefficient for out migration rate shows that the more people are leaving, the lower the fertility will be, while the positive coefficient associated with the return migration rate show that, in the counties where more migrants are coming back, fertility is higher. People avoid having children when they are in a bad financial situation. The coefficient for the unemployment rate shows that the higher the rate of unemployment, the lower the fertility. However, many returning migrants are in a good financial situation, so they can afford to have children.

33 It is possible that some of the migrants have kids while abroad, but I do not have any data regarding the number of children born abroad. 123

124

3.4. Conclusions

Given that Romania is a country with a high rate of inflation and unemployment, where almost 50% of people are below poverty and the average salary is around 150 euros, neoclassical economic theory of migration seems to explain well the reasons for which drive Romanians to look for jobs in Western Europe. There is a large gap between the incomes one can get in Romania and in Western Europe, on one hand, and, on the other hand, there are many jobs available in the Western Europe, especially in agriculture and construction industry. However, the direction of migration waves- in which countries go the emigrants from Eastern European countries- is not well predicted by the neoclassical theory. Migrants go in countries where they can adapt easily – as the case with the ethnic Hungarians migrating to Hungary, ethnic Germans going to Germany,

Romanians going to Italy and Spain and ethnic Turks from Bulgaria going to Turkey show.

The data discussed in this chapter show that migration influences other demographic phenomena, namely fertility. A high rate of out migration has a disruptive effect on fertility, on one hand, and bias fertility measures, on the other hand, because the denominator of various fertility rates is artificially inflated by including those people who are not actually in the country. Migrants tend to be young people, and they postpone having children until they achieve a better economic situation. Once they achieve that, they have children, as the results regarding return migration show: those counties with higher rate of return migration also have higher fertility. Therefore, even if on the short

125 run migration decreases fertility through disruption and biases fertility measures, in the long term it has a positive influence.

Of course, as census data show, migration is not the main factor depressing fertility – even if nobody had migrated after 1990 from Romania, fertility would not be more than

8% higher than it is. As more and more return migrants go back, fertility has a chance of increasing, especially in those areas, such as Moldavia, where the idea of having many children is still strong.

126

4. Conclusions

Why fertility declined suddenly after 1990 in Eastern Europe remains an open question, although, as I discussed in the sections above, several explanations have been given during the last years. One thing that needs to be mentioned is that the ‘sudden’ change so hotly debated is actually not that sudden for all Eastern European countries. In fact, all these countries except Poland, Romania and the former GDR, experienced a continuous rather than ‘sudden’ change in fertility after 1990 (Table 4.1).

Table 4.1. TFR for selected Eastern European countries

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

Bulgaria 1.98 2.02 1.97 1.97 1.87 1.78 1.66 1.55 1.46 1.37 1.23 1.23 1.09 1.11 1.3 1.27 1.24 1.21 Czech 1.961.941.911.941.881.91.861.721.671.441.281.181.171.161.131.141.141.16 Republic Hungary1.851.841.821.811.821.871.881.781.691.651.581.461.381.331.261.291.281.29

Poland 2.33 2.22 2.15 2.13 2.08 2.04 2.05 1.93 1.85 1.79 1.61 1.58 1.51 1.43 1.39 1.39 1.37 1.36 Romania 2.31 2.4 2.39 2.31 2.2 1.84 1.58 1.5 1.44 1.41 1.34 1.3 1.32 1.32 1.32 1.32 1.33 1.34 Slovak 2.262.22.152.152.082.092.051.931.871.671.521.471.431.381.321.281.281.29 Republic

source*: US census bureau (http://www.census.gov/cgi-bin/ipc/idbsprd) United Nations Economic Commission for Europe, Gender Statistics Database (www.unece.org) Recent Demographic Developments in Europe, 1999 *Sometimes the TFR has different values in the three databases used here. The differences are not very large (usually less than 0.1); as a rule, I used the same source of statistics for all countries in one year.

As Table 4.2 shows, if we take into consideration the proportional change in the

TFR during 1985-2002, there is not much variation in the amount of change immediately before and after 1990.

127

Table 4.2. Proportional change for TFR, 1985-2002

1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

Bulgaria 0.02 -0.03 0.00 -0.05 -0.05 -0.07 -0.07 -0.06 -0.07 -0.11 0.00 -0.13 0.02 0.15 -0.02 -0.02 -0.02 Czech -0.01 -0.02 0.02 -0.03 0.01 -0.02 -0.08 -0.03 -0.16 -0.13 -0.08 -0.01 -0.01 -0.03 0.01 0.00 0.02 Republic Hungary -0.01 -0.01 -0.01 0.01 0.03 0.01 -0.06 -0.05 -0.02 -0.04 -0.08 -0.06 -0.04 -0.06 0.02 -0.01 0.01

Poland -0.05 -0.03 -0.01 -0.02 -0.02 0.00 -0.06 -0.04 -0.03 -0.11 -0.02 -0.05 -0.06 -0.03 0.00 -0.01 -0.01 Romania 0.04 0.00 -0.03 -0.05 -0.20 -0.16 -0.05 -0.04 -0.02 -0.05 -0.03 0.02 0.00 0.00 0.00 0.01 0.01 Slovak -0.03 -0.02 0.00 -0.03 0.00 -0.02 -0.06 -0.03 -0.12 -0.10 -0.03 -0.03 -0.04 -0.05 -0.03 0.00 0.01 Republic

source: US census bureau (http://www.census.gov/cgi-bin/ipc/idbsprd) United Nations Economic Commission for Europe, Gender Statistics Database (www.unece.org) Recent Demographic Developments in Europe, 1999 and authors’ computations

Romania recorded the biggest change in fertility immediately after 1990: from above replacement level in 1989 (2.2), the total fertility rate decreased almost 20% in

1990 alone, and another 16% in 1991. As I argue in Chapter 2 (The Politics of Fertility and Their Effects in Romania), this sudden decrease is a result of changing fertility policies. Before 1989, legal abortion was hard to get in Romania and there was almost no contraceptive education or contraceptive devices. Fertility was maintained above replacement level before 1989 because of these legal barriers to avoiding pregnancies. It was a level artificially maintained. When the abortion law changed, and women had access to abortion and contraception, fertility fell to levels similar to those recorded in other Eastern European countries. While the post 1991 developments in fertility can be explained as a result of other phenomena (poverty, out-migration, changes in values and preferences regarding family), the 1989-1991 sudden fall is actually a return of fertility

128 to, let’s say, ‘natural’, ‘expected’ levels, back from the artificially-maintained above replacement fertility rate.

The high rate of out-migration from several Eastern European countries (Bulgaria,

Romania, Poland, Albania) is seen by some researchers as a powerful factor in explaining the declining fertility of the region. Chapter 3 (Influences of migration on fertility) analyzes the importance of out migration for fertility decline in Romania. As census and survey data show, not only are the fertility measures biased because of the high rate of emigration, but out-migration also has a disruptive effect on fertility, as migrants postpone having children until they get the financial means to establish a family.

However, as more migrants return home, fertility improves, so in the long term fertility can benefit from emigration. Changes in ethnic composition also have a positive effect on fertility, as the new generations have a higher percentage of Gypsies, a minority with high fertility. As the data from Table 4.1 shows, fertility began to increase slightly after

2000, and this can be seen as the result of an increasing percentage of Gypsies in the population at reproductive age.

In conclusion, in this dissertation I argue that the influence of five factors should be taken into account when fertility decline in Eastern Europe is studied. Changes in attitudes regarding family size are a phenomenon that began a long time ago – as the case of Romania shows, as early as the 1960s, women wanted to have less than two children.

However, these changes in attitudes did not have an immediate effect on fertility in

Romania during 1967-1989 because of the very restrictive access to contraceptives and abortion. When the means of limiting fertility became available – as happened after 1989

– the number of children born declined sharply. The most important factor explaining the

129 rapid decline of fertility during 1989-1992 – from 2.2 to 1.5 - is the change in the legal constraints regarding abortion. In 1990, the first year when abortion was easily accessible, there were more than 3 abortions for each live birth, which shows both women’s lack of contraceptive education and a strong desire to limit the number of children.

While changes in the legal restrictions regarding abortion in 1989 can be shown to influence fertility decline in the short term - basically during 1989-1992 - changes in women’s educational attainment, women’s labor force participation, the ethnic composition of population and the rate of urbanization played a long term role in fertility decline in Romania. After 1945, communist governments in Eastern Europe invested many resources in increasing women’s status in society by encouraging women to work and to get educated. As a result, fertility declined, despite the various pronatalist policies.

Changes in the ethnic composition of population also played a role in the fertility changes. As the Romanian case shows, some ethnic groups have high fertility, while other ethnic groups have lower fertility, which leads to a changing ethnic composition of the population. Fertility is lower when the proportion of low fertility ethnic groups is higher, and is higher when the proportion of high fertility ethnic groups increases. High levels of urbanization recorded by the Eastern European countries after 1945, and in

Romania in particular, also played a role in declining rates of fertility. Urban populations have lower fertility rates than rural populations have, so an increasing proportion of urban residents in Eastern Europe resulted in lower fertility at the national level.

After 1990, some of the Eastern European countries recorded high levels of out- migration. Emigration is an important factor in declining fertility, especially for the

130 second part of 1990, when the stock of emigrants increased dramatically. It influences fertility in two different ways. On one hand, migrants are usually young people who postpone having children and families and go to work abroad. As I showed in Chapter 3, without migration, fertility in Romania would be 5% higher during 1992-2002. On the other hand, fertility measures are biased by emigration because the denominator of the

TFR and of other measures is inflated with a population that is not actually at risk of having children in the country. It was shown that, for the Romanian case, total fertility rates would be 8% higher if migration were taken into account.

There is no society that has succeeded in increasing fertility from below replacement to above replacement through positive, non-coercive policies. Aging societies, such as Eastern European began to be, had to find means other than fertility policies to fight the increasing percentage of retired individuals. One way is to increase the age at retirement, as Romania has tried to do. Another way is to encourage immigration, as Czech Republic began to do. None of these ways is risk free; in Italy, people protested strongly against changing the age at retirement, as nobody is happy with the idea of working longer that they expected.

This research argues for a more careful analysis of fertility decline in Eastern

Europe after 1990. As Hirschman argues for the case of demographic transition theory, theories need more ‘flesh’ when applied to specific settings. General theories – as second demographic transition theory or economic explanations – developed initially to explain the decrease in fertility in Western Europe and other industrialized nations, should be more carefully applied when countries of Eastern Europe are the object of study. As the

Romanian case shows, factors such as population policies, the presence of ethnic

131 minorities with different fertility patterns, and out-migration play an important role in explaining fertility changes in some former communist countries, while for other countries they might be insignificant. History matters in Eastern Europe, probably because these countries have had such complicated political and economic paths; it is not only the immediate history that matters, but also the earlier one, that is, before 1945

132

Bibliography

***. 1969. Romanian Census, 1966, Central Department for Statistics, Romania

***.1980. Romanian Census, 1977, Romanian Commission for Statistics

***. 1994, Romanian Census, 1992, Romanian Commission for Statistics

***. 2003. Romanian Census, 2002. Romanian Institute for Statistics

***.2001. Romanian Demographic Yearbook. Romanian Institute for Statistics

Achim, Viorel.1998. Tiganii in istoria Romaniei [Gypsies in the Romanian History], Bucharest : Ed. Enciclopedica, 1998

Althaus, F. 1992. “Replacement –Level Fertility has Now Become the Rule in Most of Eastern Europe”. Family Planning Perspectives, 24

Avdeev, A. 2001. “The extent of the fertility decline in Russia: is the one-child family here to stay?” , paper presented at the IUSSP Seminar on International Perspectives on Low Fertility: trends, theories and policies, Tokyo

Baban, Adriana. 1999. “Romania”, in From Abortion to Contraception, edited by Henry P. David and Joanna Skilogianis, London: Greenwood Press

Barany, Zoltan. 2002. The East European Gypsies. Regime Change, Marginality, and Ethno politics. Cambridge: University Press

Berelson, Bernard. 1974. “An Evaluation of the Effects of Population Control Programs”, Studies in Family Planning,vol.5 (1), pp.2-12

Bhrolchain, M.N. 1992. "Period Paramount? A Critique of the Cohort Approach to Fertility." Population and Development Review 18(4):599-629

Bongaarts, John and Griffith Feeney. 1998. “On the quantum and tempo of fertility”, Population and Development Review 24 (2): 271-291.

Bongaarts, John and Griffith Feeney. 2000. “On the quantum and tempo of fertility: Reply”, Population and Development Review 26 (3): 560-564.

Botev, Nikolai. 1990. “Nuptiality in the Course of the demographic Transition: The Experience of the Balkan Countries”, Population Studies, 44

133

Braun, Lily. 1987. Selected writings on feminism and socialism (translated and edited by A.G. Meyer). Bloomington: Indiana University Press

Bucur, M. 2002. Eugenics and Modernization in Interwar Romania, 2002, Pittsburgh: University of Pittsburgh Press

Caldwell, John C. and Thomas Schindlmayr. 2003. “Explanations of the fertility crisis in modern societies. A search for commonalities”, Population Studies, vol.57, no.3, pp.241- 263

Carlson, Elwood, Meguni Omori. 1998. “Fertility Regulations in a Declining State Socialist Economy: Bulgaria, 1976-1995”. International Family Planning Perspectives, 24

Chesnais, Jean-Claude. 1992. The Demographic Transition. Stages, Patterns, and Economic Implications. A Longitudinal Study of Sixty-Seven Countries Covering the Period 1720-1984. trasl. by Elisabeth and Philip Kreager. Oxford: Oxford University Press

Coale, A.J. and S.C. Watkins (ed.). 1986. The Decline of Fertility in Europe. The Revised Proceedings of a Conference on the Princeton European Fertility Project, Preface and chapters 2, 4, 7, 8, 10, 11. Princeton: Princeton University Press

Coleman, David A. 1992. “Does Europe Need Immigrants? Population and Work Force Projections”. International Migration Review, 26, 2, 413-461

Costa-Foru, Xenia. 1945. Cercetarea monografica a familiei [Monographic research methodology of family]. Bucuresti: Fundatia Regele Mihai I

Comisia Nationala de Demografie (CND). 1980. Ancheta Fertilităţii Populaţiei Feminine din R.S.România. Anul 1978. [ Fertility Survey of Women in Socialist Republic of Romania. Year 1978], Consiliul de Stat, RSR

David, Henry. 1999. “Hungary”, in From Abortion to Contraception, edited by Henry P. David and Joanna Skilogianis, London: Greenwood Press

Davis, Kingsley. 1963. “The Theory of Change and Response in Modern Demographic History”. Population Index, 29, 345-366

Demeny, Paul. 1986. “Pronatalist Policies in Low-Fertility Countries: Patterns, Performences, and Prospects”, in Population and Development Review, 12, Issue Suplement: Below Replacement Fertility in Industrial Societies: Causes, Consequences, Policies, 335-358

134

Dyami, Amilda & Pamela Pine. 1999. “Albania”, in From Abortion to Contraception, edited by Henry P. David and Joanna Skilogianis, London: Greenwood Press

Ebanks, G.E, P.M. George, C.E. Nobbe. 1975. “Emigration and Fertility Decline: the Case of Barbados”, Demography, 12, 3, 431-445

Eberstadt, Nicholas. 1994. “Demographic Shocks after Communism: Eastern Germany 1989-1993”, Population and Development Review, 20, 1

Engels, Friedrich. 1978 [1884]. “The Origin of the Family, Private Property, and the State”, in R. Tucker. The Marx-Engels Reader. New York: WW.Norton&Company.

Espenshade, T.J. 1994. “Does the Treat of Border Apprehension Deter Undocumented US Immigration?”. Population and Development Review, 20 (4), 871-92

Fassman, H.&R.Munz. 1992. “Patterns of International Migration in Western Europe”. Population and Development Review, 18, 3, 457-480. Fassman, H.&R.Munz. 1994. “European East-West Migration. 1945-1992”. International Migration Review, 28, 3, 540-538. Falkingham, J. A.Gjonca. 2001. “Fertility transition in Communist Albania, 1950-90”. Population Studies, 55, 309-318

Fialova, L., Z. Pavlik, P.Veres.1990.”Fertility Decline in Czechoslovakia during the Last Two Centuries”. Population Studies, 44,1

Frątczak, Ewa & Aneta Ptak–Chmielewska. 2001. “Fertility and Family Life Cycle Changes in Poland and the Second Demographic Transition”, presented at the EURESCO Conference “The second demographic transition in Europe”, Bad Herrenalb, Germany, 23-28 June 2001

Fratczak, Ewa, Margarete Kulik, Marcin Malinowski. 2003.” Legal Regulations related to Demographic Events and Processes: Social Policy Pertaining to Children and family- Poland, Selected years 1945-2003”. European Population Conference in Warsaw, Aug. 26-30

Freeman, G.P. 1992. “Migration Policy and Politics in the Receiving States”, International Migration Review, 26, 4, 1144-1167

Gauthier, Anne Helene & Jan Hatzius. 1997. “Family Benefits and Fertility: An Econometric Analysis”, Population Studies, 51, 3

Georgescu, D.C.1940. La fertilité différentielle en Roumanie [Diffeential fertility in Romania]. Bucureşti: Institutul Central de Statistică

135

Ghetau, V. 1997. Evolutia fertilitatii in Romania. De la transversal la longitudinal [The Fertility Evolution in Romania. From transversal to longitudinal perspective] Bucuresti:CIDE

Goody, Jack. 1996. “Comparing Family Systems in Europe and Asia: Are There Different Sets of Rules?”, Population and Development Review, 22,1

Goldstein, S., A. Goldstein. 1981. “The Impact of Migration on Fertility: an ‘Own Children’ Analysis for Thailand”, Population Studies, 35, 2, 265-284

Hammar, Tomas and K.Tamas. 1997. “Why Do People Go or Stay?” in T.Hammar & all (ed.). International Migration, Immobility and Development, Oxford: Berg

Henshaw, S, S. Singh, T. Haas. 1999. “The Incidence of Abortion Worldwide”. International Family Planning and Perspectives. Supplement, 25

Hirschman, Charles. 1994. “Why Fertility Changes”. Ann.Rev.Sociol. 1994. 20:203-233 Hitchins, Keith. 1994. Rumania 1866-1947. Oxford: Clarendon Press

Hajnal, John. 1982. “Two Kinds of Preindustrial Household Formation System”. Population and Development Review, 8. No.3.

Hervitz, Hugo M.1985. “Selectivity, Adaptation, or Disruption? A Comparison of Alternative Hypotheses on the Effects on Fertility: The Case of Brazil”, International Migration Review, 19, 2, 293-317

Ilea, Th. P.Mureşan, I.Ionescu, A.Pavel, M.Bilegan, M.Teodorescu. 1969. Cercetări medico-sociale asupra fertilităţii din R.S.România [Socio-medical Studies of Fertility in R.S. Romania]. Ministry of Health van Imhoff, Evert and N.Keilman. 2000. “On the quatum and tempo of fertility: Comment”, Population and Development Review 26 (3): 549-553 Jandl, Michael. 2003.” Moldova Seeks Stability amid Mass Emigration”, Migration Information Source (www.migrationsource.org)

Jessop, Bob. 1998. “Karl Marx”, in Rob Stones, Key Sociological Thinkers. New York: New York University Press.

Jonsson, Stefan Hrafn & M.Rendall. 2004. “Fertility Contribution of Mexican Immigration to the United States”, Demography, 41,1,129-150

Kim, Young and Robert Schoen. 2000. “On the quatum and tempo of fertility: Limits to the Bongaarts-Feeney Adjustment”, Population and Development Review 26 (3): 554-559

136

Kulcsar, Laszlo. 2003. “Demographic trends in Hungary since WWII: what did change after the fall of socialism?”, manuscript

Lenin, V.I. 1977. Karl Marx and His Teaching. Moscow: Progress Publishers

Lindstrom, Davis P., S.G.Saucedo. 2002. “The Short and Long –Term Effects of US Migration Experience on Mexican Women’ Fertility”, Social Forces, 80.4, 1341-1368

Martin, P. and E. Midgley. 1999. Immigration to the United States. Population Bulletin, Vol.54, 2

Manfras, K. 1992. “Europe: South-North or East-West Migration?”, International Migration Review, 26, 2, 388-400.

Massey, D. and K.E. Espinosa. 1997. “What’s Driving Mexico-U.S. Migration? A Theoretical, Empirical, and Policy Analysis”. American Journal of Sociology 102: 939- 999

Massey, D, J.Arango, G.Hugo, 1993.”Theories of International Migration: A Review and A.Kuouaouci, A.Pellegrino, Appraisal”. Population and Development Review J. E. Taylor. 19:431-466.

Maddison, Angus. 1995. Monitoring the World Economy, 1820-1992. Paris: OECD

Marx, Karl. 1978 [1848]. “Manifesto of the Communist Party”, in Key Sociological Thinkers. New York: New York University Press

McDonald, Peter. 2000. “Gender Equity in Theories of Fertility Transition”. Population and Development Review 26 (3):427-439

Metes, Stefan. 1977. Emigrări româneşti din Transilvania în secolele XII-XX.[ Romanian migrations from Transylvania between 13th –20th centuries]. Bucureşti: Editura Ştiinţifică şi Enciclopedică

Milosz, Czeslaw.1964.Une autre Europe. Gallimard

Muresan, Cornelia. 1996. “L'evolution demographic en Roumanie. Tendances passees (1948-1994) et perspectives d'avenir (1995-2030)” [ Demographic evolution in Romania. Passed tendencies (1948-1994) and future prospective (1995-2030)], Population, 4-5

Mureşan, P., I.Caniola, I.Copil, E.Malceolu, R.Paraschivescu, B.Pascu, C.Prisăcaru, L.Roynatovski, D.Segal. 1977. Studiu longitudinal al fertilităţii în R.S. România [ Longitudinal Study of Fertility in R.S. Romania]. Ministry of Health, Centre for Computations and Health Statistics

137

De Nevers, Renee. 2003. Comrades no more. The seeds of Change in Eastern Europe. Cambridge: The MIT Press

Potot, Swanie. 2000. ”Mobilités en Europe. Étude de deux réseaux migratoires roumains [Mobility in Europe. A Study of Two Romanian Migration Network], Sociologie Romaneasca, 2

Philipov, Dimiter.2001.“Low fertility in Central and Eastern Europe: Culture or economy?”, paper presented at the IUSSP Seminar on International Perspectives on Low Fertility: trends, theories and policies, Tokyo, March 21-23, 2001, USSP Working Group on Low Fertility and National Institute for Population and Social Security Research, Japan

Ranjan, Priya. 1999. “Fertility Behaviour under Income Uncertainty”, European Journal of Population, 15, 25-43

Le Rider, Jacques. 1994. Mitteleurope. Presses Universitaires de France

Root, B.D. & G.F. de Jong. 1991. “Family Migration in a Developing Country”, Population Studies, 45, 221-233

Sandu, D. 2001. ”Migratia Transnationala a Romanilor din Perspectiva unui Recensamant Comunitar” [International Migration of Romanians studied with Community Census ], Sociologie Romaneasca, 3-4

Serbu, G.R. 2000. “Evoluţia fertilităţii populaţiei feminine din R.P.R. în perioada 1900- 1960” [Fertility evolution in Romania between 1900-1960]. (republished in) Populaţie şi societate, 4-5

Stahl, Paul H. 2000. Triburi si Sate din Sud –Estul Europei.[ Tribes and Villages in South Eastern Europe]. Bucuresti: Paideia

Sklar, June. 1974. “The Role of Marriage Behavior in the Demographic Transition: The Case of Eastern Europe around 1900”. Population Studies, 28, 2

Sobotka, Tomáš, Kryštof Zeman & Vladimíra Kantorová. 2001.’ Second demographic transition in the Czech Republic: Stages, specific features and underlying factors’, presented at the EURESCO Conference “The second demographic transition in Europe”, Bad Herrenalb, Germany, 23-28 June 2001

Shevstova, Lila. 1992. “Post-Soviet Emigration Today and Tomorrow”. International Migration Review, 26,2 , 241-257

Singley, Susan & Nancy Landale. 1998. “Incorporating Origin and Process in Migration- Fertility Frameworks: The Case of Puerto Rican Women”, Social Forces, 4, 1437 - 1464

138

Stark, O.& J.Taylor. 1989. “Relative Deprivation and International Migration”, Demography, 26, 1-14

Straubhaar, Th. 1986. “The Causes of International Labor Migration – A Demand – Determined Approach”. International Migration Review, 20, 4, 835-855

Teitelbaum, M.S.1972. “Fertility Effects of the Abolition of Legal Abortion in Romania.” Population Studies, 26(3)

Todorova, Maria N. 1993. Balkan Family Structure and the European Pattern. Demographic developments in Ottoman Bulgaria. Washington: American University Press

Todorova, Maria N. 1997. Imagining the Balkans. Oxford: Oxford University Press Wilson, Chris and P.Airey. 1999. “How can a homeostatic perspective enhance demographic transition theory?”, Population Studies, 53, 117-128

Titkow, Anna. 1999. “Poland”, in From Abortion to Contraception, edited by Henry P. David and Joanna Skilogianis, London: Greenwood Press

Tomka, Béla.2002. “Demographic Diversity and Convergence in Europe, 1918-1990: The Hungarian case”, Demographic Research, vol.6, (www.demographic-research.org)

Vassiliev, Dmiter. 1999. “Bulgaria”, in From Abortion to Contraception, edited by Henry P. David and Joanna Skilogianis, London: Greenwood Press

Westoff, Charles F.& Norman B. Ryder. 1977. “The Predictive Validity of Reproductive Intentions”, Demography, 14, 4

Wynnyczuk, Vladimir & Radim Uzel. 1999. “Czech Republic and Slovak Republic”, in From Abortion to Contraception, edited by Henry P. David and Joanna Skilogianis, London: Greenwood Press

Zolberg, Aristide. 1989. “The New Waves: Migration Theory for a Changing World”. International Migration Review 23: 457 – 485

139

VITA

Cristina Bradatan

Education: 2004: PhD, Sociology & Demography, Pennsylvania State University, Minor: Quantitative Methods 1999-2000: Postgraduate studies in Demography, Charles University, Prague 1999: MA in Sociology, University of Bucharest 1998: BA in Sociology, University of Bucharest 1997: MS in Applied Statistics, University of Bucharest 1996: BS in Mathematics, University of Bucharest

Academic and professional experience: 2002-2004: research assistant, Department of Demography, Pennsylvania State University summer 2003: instructor, Department of Sociology, Pennsylvania State University 2001-2002: research assistant, Department of Sociology, Pennsylvania State University 2000-2001: Fellow of Graduate School, Pennsylvania State University 1999-2000: part time consultant, Department of Actuaries, Nationale Nederlanden Life Insurance, Prague, Czech Republic 1998–2000: researcher, Center for Population Studies, National Academy of Romania 1996-1999: assistant lecturer, Department of Applied Mathematics and Computer Science, University of Bucharest

Research and teaching interests • Migration and family changes of Europe • Statistical methods and demographic techniques • Psychoanalysis and society • Sociology of food

Publications (unless otherwise specified, all publications are in English)

In Academic Journals: - “Cuisine and cultural identity in the Balkans”, The Anthropology of Eastern Europe Review, vol.21, no.1 (spring 2003), 43-49 - “Family systems in Eastern Europe”, Romanian Sociology, no.1-4, 2001, 366-374 [in Romanian] - “On Vladimir Trebici. The Demographer”, Population & Society, no.1, 2000 [in Romanian] - “Suicide as a Social Phenomenon”, Romanian Sociology, no.2, 1999, 79-90 [in Romanian] - “On Some Population Problems in Romania”, Population & Society, no.4, 1998 [in Romanian]

In Collective Volumes: - “Hispanic Families in the United States: Family Structure and Process in an Era of Family Change” (with N. Landale and R.S. Oropesa) (to appear) - “Differential Mortality and Regional Development”, CIDE, Bucharest, 2000 [in Romanian] - “Suicide and Aggression” in Social Conditioning of the Mortality, CIDE, Bucharest, 1999 [in Romanian]