Note: Where available, the PDF/Word icon below is provided to view the complete and fully formatted document
Low fertility in Central and Eastern Europe: culture or economy? Paper to be presented at the IUSSP Seminar on International Perspectives on Low Fertility: Trends, Theories and Policies, 21-23 March 2001



Download PDFDownload PDF

- 1 - 09/06/02

Low fertility in Central and Eastern Europe: Culture or economy?

Dimiter Philipov

Max-Planck Institute for Demographic Research

Rostock, Germany

philipov@demogr.mpg.de

FIRST DRAFT

Paper to be presented at the IUSSP Seminar on "International Perspectives on

Low Fertility: trends, theories and policies", Tokyo, March 21-23, 2001.

IUSSP Working Group on Low Fertility and National Institute for Population and

Social Security Research, Japan.

- 2 - 09/06/02

Introduction

The socialist totalitarian regime in the East European countries is gone for 11 completed years, a period known as one of transition from a planned to a market society. The planned society collapsed by itself at the very start, leaving place for a market-based social order that was far from orderliness. At the beginning of the transition people hoped that disorderliness would last a couple of years. It took longer. Nowadays some countries are close to the end of the transition, while others have still a long way to go. The transition relates to drastic changes in all spheres of life. The state was entirely reorganized and the economy was restructured on the basis of rising private property and free market relations. Some people came to know impoverishment, others enrichment. During the decade the countries from the region diverged considerably with respect to their move towards a new society.

The total change in life did not overpass demographic events. Marriages and births decreased and were postponed, the share of non-marital unions and extra-marital births increased. Some of the demographic changes, such as the abrupt fall in fertility, were unprecedented in the history of mankind in peaceful times. The total fertility rate (TFR) decreased to app. 1 in at least several countries from the region. These trends developed in all countries in the region. Thus the populations in the region experience a demographic convergence during times of political, social and economic divergence.

Demographers that study these events explain them in diverse ways. Some find the economic difficulties of the transition as the primary cause for the new demographic trends. Indeed, the pervasive economic decline has brought about a considerable increase in the direct costs of children. Others argue that the drop in fertility is basically the result of long-standing cultural changes who have been developing before the start of the transition. The latter has only contributed to their rise at the surface and acceleration. Both approaches give a valuable insight to the understanding of recent fertility changes in the countries from the region.

This paper argues that both approaches are difficult to disentangle because of the lack of data. In addition they do not incorporate in full the quintessence of the transition, describing it in too general terms. The transition is a period of discontinuation, characterized by a sudden break of institutions and social order and their gradual replacement by new ones. It affects norms, values, preferences, attitudes, and behavior. The latter may break and give rise to disorderliness and anomie. Uncertainty rises in such a state of disorderliness. Affected people may decide to postpone and even reject crucial and irreversible life events, such as a birth of a child. People react to uncertainty in different ways. Some may remain passive, while others undertake diverse actions, or coping strategies. Persons who manage to develop successful coping strategies may lower their burden on uncertainty and hence be less willing to postpone or reject a birth.

The first section introduces the reader to the turbulent social and economic development in the region. The next one describes the recent fertility trends in the Central and Easter European (CEE) countries during the 90-s. The third and fourth sections discuss the two grand explanatory approaches to that fall. In the fifth section we emphasize discontinuity as an essential characteristic of the transition, and the disorderliness and anomie in society it has caused.

- 3 - 09/06/02

The lack of data is a significant obstacle for an in-depth comparative research and assessment of theoretical approaches. The paper relies basically on macro-level statistics data. Use is made also of individual level data from the FFS and the EVS which contain only partial information. Due to the data limitations the question posed in the title remains without a definitive answer.

The paper discusses fertility change in 15 countries, distributed in four regions (see also map on figure 1). (a) Central European countries: Czech Republic, Hungary, Poland, Slovakia, plus Slovenia.

(b) Baltic countries: Estonia, Latvia, Lithuania. (c) South-Eastern European countries: Bulgaria, Macedonia, Romania. (d) CIS countries (Commonwealth of Independent States): Belarus, Moldova, Russia, Ukraine. The discussion excludes other ex-Yugoslav countries that have experienced wars, as well as Albania and the Trans-Kaukazian countries where the social aftermath of the transition may have affected in a specific way the demographic trends; moreover their demographic patterns are specific and call for a separate discussion.

1. The context of the transition

The totalitarian regimes collapsed in the Central European ex-socialist countries in the end of 1989, and the USSR disintegrated in the end of 1991. These are the starting years of the transition. The description given below refers more or less to all countries in the region. The major differences among the countries are in the pace, timing, and intensity of these historical events. We first focus on the common features of the transition.

1.1. Political and economic aspects of the transition Politically, the transition started when a multi-party ruling system replaced virtually within days the totalitarian regime. The civil society began its gradual revival. The ex-communist parties retained legally the power, or returned to power for some periods during the decade. Legally, the transition started with significant constitutional changes, or even adoption of new constitutions, with the purpose to abolish the legal power of the communist ideology. The path towards a market economy required the change of numerous laws and acts. The speedy legal changes called for repeated improvements of the new laws. Thus the legal system is in a process of continuous renovation. The transition in the economic system took the course from a planned-oriented to a free market economy. The reforms began right at the start with liberalization of prices, trade, markets and market entry. The private sector marked a significant rise, due by and large to privatization and reconstitution of old property in some of the countries. The collapse of Comecon and the USSR demolished traditional secure markets. Enterprises re-oriented their trade, with the preference being towards the more developed countries in Western Europe. The financial markets revived with the birth of numerous banks. Stock markets revived. Currency convertibility was eased. Liberalization was shortly followed

- 4 - 09/06/02

by macroeconomic stabilization. The latter had to fight rising inflation through fiscal, monetary, and wage policies, although with diverse success. The reforms started at a time when the economies of all CEE countries were near collapse. The initial conditions were substantially different: some countries inherited significant foreign debts, others had their export oriented exclusively towards Comecon or the ex-USSR. Nevertheless, all countries experienced an economic shock with the start of the reforms. This shock is evident in the drop of the GDP (table 1), which decreased by at least 10 points in only 1-2 years. The further path of the change in the GDP is diverse. In some countries, notably the Czech Republic, Hungary, Poland, and Slovenia, it gradually recovered or even surpassed its 1989 level towards the end of the decade. In the Baltic countries the recovery is at a slower pace but is evident, while in the rest of the countries, notably those in South-East Europe and the European CIS countries, the recession continues. The rate of change in the GDP corresponds to the pace, timing, and intensity of the reforms. The better-off countries initiated rapid adjustment reforms, often close to laissez-faire, while others pertained to sustained change, or gradual economic adjustment.

1.2. Social aspects of the transition The social aspects of the transition relate mostly to changes in educational enrolment, social stratification and income inequality, as well as rise in impoverishment and unemployment. They bear a significant impact on fertility change and for this purpose are discussed in some detail. Education. The liberalization process, both economic and political, caused the rise of numerous new opportunities for personal choice. One such direction of primary importance is the spread of new professions and occupations. They relate essentially to the demand side of the newly established free markets. For example, business behavior and entrepreneurial attitudes unfolded widely, as well as occupations linked to the financial market and the legal system. Business management, accounting and finances, banking, law, came to be among the fashionable and preferred professions. The supply side of this labor market was poor and hence gave impetus to the extend of proper education. Liberalization made it possible for private schools and universities to appear. Thus the conditions for the increase in education, both extensively: among an increasing number of people, and intensively: through its prolongation, were at hand. Statistical data are insufficient to illustrate this change in detail, but a few examples are available. Macura (2000, page 199) notes that the proportions of 18-22 year old women in Hungary, Poland, Bulgaria, and Romania, attending higher education institutions have doubled in 1997 relative to 1989. An increase in these proportions although not as high was noted in the other CEE countries. The 1997 enrolment rates were around 15-25 per cent, which is considerably lower than that observed in Western Europe, hence the extend of education is likely to continue. Frątczak (2000, p.32) notes that the enrolment ratio for universities for persons aged 18-24 increased from 6% in 1990 to 25% in 1996. The rise in educational expenses forced some students to combine work with study, possibly breaking the study for some time. Such effects contribute to a postponement of education. Unemployment. Unemployment did not exist, at least officially, during the totalitarian regime, when even frictional unemployment was not registered. Table 2 indicates a considerable rise in unemployment during the 10-year period. Data not

- 5 - 09/06/02

presented here reveal well-known patterns of unemployment among groups of the population, such as higher unemployment among women and young adults. A specific feature of unemployment in the most of the CEE countries is the long-term unemployment and the associated high level of discouraged workers.

Impoverishment. It is another new aspect of social life in the CEE countries. Impoverishment spread quickly and reached considerable levels in some countries. Table 3 presents the percentage of the population below a poverty line of 4$ per capita per day. Like unemployment, poverty starts from very low and rises to very high levels.

Income inequality. Inequality existed during the totalitarian regime at least because of the existing of nomenclatura. The free market regime brought about the spread of both impoverishment and enrichment. The Gini coefficient for some CEE countries is given in table 4.

Declining effectiveness of family and other policies. Macura (2000) provides a concise discussion on family policies. He points that at the start of the transition all countries had well designed and rich family policies with effective instruments, such as maternity and childcare leaves, compensation for income during maternity leave, child allowances. While leaves have been preserved, payments have decreased, in some countries considerably due to periodic and high inflation. The number of kindergartens decreased, partially due to privatization of housing stock

The transition had its specific trajectory in each country. In some of the countries the initial economic recession was soon followed by economic recovery. These are the countries from the Central European region (a), as denoted in the end of the introduction. The Baltic countries, our region (b), have speeded their development and are catching up group (a). The countries from group (c) have a longer way to go, although in some respects they are close to the Baltic countries. Finally, the CIS countries move along a slowest pace. The comparison is based on the assumption that all countries have a common aim, namely a democratic society of a Western European type. At the beginning of the transition the countries were considerably closer to each other where social-economic indicators are considered. Hence this was a period of divergence in this part of Europe.

2. Fertility changes - general description

This chapter provides a brief description of the fertility changes before and during the transition period. The demographic literature suggests a number of publications that provide such a description, see for example Monnier (1996), Monnier and Rychtarikova (1992), Kucera et al. (2000), Macura (1999 and 2000), as well as the FFS CRS-s1.

Pre-transition fertility. The period total fertility rate (TFR) marked significant fluctuations during the decades preceding the 90s. Table 5 gives its values for selected years. Its tendency towards a decrease was countered by pronatalist population policies.

1 Fertility and Family Survey Country Report Series, issued by the Population Activities Unit at the Economic Commission for Europe, United Nations, Geneva. SCR-s are available for Poland, Hungary, Estonia, Latvia, and will be available for Bulgaria, Lithuania, Slovenia, and possibly for the Czech Republic. These countries from Eastern Europe have carried out the FFS.

- 6 - 09/06/02

For example, the TFR went below replacement level during the middle of the 60s in Bulgaria and Romania, and a few years later in ex-Czechoslovakia, GDR, and Hungary. In all these countries this downturn was soon followed by the adoption of population policy measures, followed later by the USSR, that alleviated the fluctuations of the TFR close to replacement level. During the 80s the TFR was slightly above replacement level in some countries (notably in Poland, Romania, nowadays Yugoslavia - the latter is not included in the table), slightly below it in others (like Bulgaria, Czech Republic, Hungary), and around replacement in nearly all others. Important exceptions are Slovenia, Croatia and the GDR, where the fall was below 1.8 towards 1989.

The period first-order TFR (denoted by TFR1; not included in this paper) was close to unity in all the countries from the region, with the exception of the GDR and Hungary, where it dropped towards 1989 below 0.8. TFR1 levels as high as these indicate a nearly universal birth of a first child. This pattern was widely spread over time as well. The second-order TFR2 was also very high: its values remained above 0.7. These high levels indicated the prevalence of the two-child model. The latter pattern was widely spread among the diverse countries. Order-specific completed cohort fertility supports the period-based observations, indicating even closer values to 1 for the cohort TFR1.

The mean age at birth of the first child was very low in all the countries. It remained nearly constant during the years preceding 1989, and for this latter year its values can be seen in table 6. These low values indicate to the prevalence of a pattern of early childbearing all over the region. The age-specific fertility rates data, not shown here, demonstrate that the age schedules of the birth of the first child were narrow: hence the birth of the first child usually occurred within a narrow age span. The data also reveal that fertility stopped at an early age: by age 30 around 90% of all births have already taken place.

The transition countries were much more divergent with respect to extra-marital births before the start of the transition than to any other fertility-related demographic trend. This level was higher in the Baltic countries, Russia, the GDR, and Yugoslavia than elsewhere. Towards the end of the 80-s the share of these births marked a slight increase.

Birth control was dominated by the usage of traditional methods, and particularly coitus interruptus. Modern methods became available in limited amounts during the 80s, and were available only in the Central European socialist countries.

This short outline suggests the domination of a fertility pattern characterized with: - nearly universal birth of a first child; - early start of childbearing, clustered into a narrow age interval; - prevalence of the two-child family model. This outline corresponds to a family formation pattern that abounds to the east of Hajnal’s line, characterized exactly by early and nearly universal marriage. Hajnal line does not delineate entirely all countries considered here, many of them being crossed by the line. The historically established patterns in these countries have been different and were closer to the Western European pattern. Hence the uniformity of the demographic behavior over the region cannot be explained in a historical perspective. The totalitarian regime has had its contribution to a convergence in the demographic regime.

Slovenia, Croatia, and partially other ex-Yugoslav countries are important exceptions of all or at least some of patterns mentioned above. The populations of these

- 7 - 09/06/02

countries were on the start towards demographic patterns observed in Western Europe since the beginning of the 80s. The difference can be associated to their exceptional for the region political and economic status yet before the start of the transition.

Fertility changes during the transition period. The drop in the observed period TFR (table 5) indicates a swift and considerable drop in the level of fertility. During the second half of the decade the TFR was below 1.5 in all countries. In some countries the decrease was slower during the first few years of the transition (Czech and Slovak Republic, Hungary, Poland). In nearly all other countries it was steepest namely during these first years. Notably, the decrease began in the ex-USSR countries before the collapse of the state, exception in Lithuania where it started with the achieved political independence. In Slovenia, as well as in the other ex-Yugoslav republics, the decrease was not as steep for the TFR was already low around 1989.

Towards the end of the 90s the mean age of childbearing marked a considerable increase in all countries. In order to avoid statistical problems connected to changes in the weights of different birth order the discussion focuses on women’s mean age of birth of first child (table 6). The mean ages of mothers at higher-order births followed crudely the same patterns. Changes in the latter can be divided into two periods. During the first couple of years of the transition, the mean age of birth of first child in the Central and Southern European countries usually did not change or even slightly decreased. The drop was stronger in the Slovak Republic where the mean age decreased by about 0.5 years. The same was observed in the ex-USSR countries. In all countries the initial fall was followed by a rapid rise. The latter was most notable in Croatia and Slovenia, where it was around 2 years. The increase was less significant in the ex-USSR countries because it was shorter (the transition started in these countries two years later).

A rise of the mean age of birth of the first child by 1.5 to 2 years within a period of less than 10 years indicates a break from recent behavior. In some of the countries, like the Baltic ones, this process is a return to a behavior observed several decades earlier. These countries had a traditional pattern of family formation that corresponded to the Western European type, i.e. to the west of the Hajnal line. Vikat (1994) gives more details on this subject in the case of Estonia.

It is well known in demography that postponement of births distorts the TFR where its interpretation as the average number of children is considered. Bongaarts and Feeney (1998) suggested a simple formula that allows the adjustment of the TFR for this distortion, by delineating tempo from quantum effects. Philipov and Kohler (2001) used it for the analysis of fertility change in Bulgaria, the Czech Republic, Hungary, Poland, and Russia. Figure 2 reproduces their graphs where the overall TFR is considered. The figure includes graphs for Lithuania and Slovakia, where data have become available.

Figure 2 gives the observed TFR and its adjusted counter-part, plotted on the left axis. The mean age of childbearing is plotted on the right axis. The pattern of change in the latter did not differ significantly from that of the mean age at birth of the first child (table 6). It is important to note again that this mean age increased in all countries by the end of the 90s. Countries differed only during the first few years, when the mean age of childbearing decreased in the ex-Soviet and the Southern European countries.

These changes in the mean ages caused the appearance of diverse modifications in the tempo and quantum constituents of the TFR, and therefore in the adjusted TFR.

- 8 - 09/06/02

During the first half of the 90s, the adjusted TFR remained at around replacement level in the Czech and Slovak Republic, Hungary, and Poland. The tempo effect was initially small in the Czech Republic, Poland, and Hungary, while in Slovakia, Bulgaria, Lithuania, and Russia it was negative, i.e. the adjusted TFR was lower than the observed one. Hence the adjusted TFR decreased along with the observed one in Bulgaria, Lithuania, and Russia. During the second half of the 90s the tempo effect was considerable in all the countries, and therefore the adjusted TFR was considerably higher than the observed TFR.

The patterns described above allow the generalization that the tempo effect was more or less uniform in the countries from Central Europe and therefore the fall in the TFR was due to tempo effect. In the other countries the fall in the TFR was entirely due to quantum in the beginning of the period, while the tempo effect increased during the course of the decade. These countries experienced like a “quantum shock” at the start of their fertility fall. Towards the end of the decade all countries exhibited a uniform pattern of fertility decline, as measured by the TFR and its tempo and quantum components.

Table 7 gives the values of the observed and the adjusted tempo-free TFR1. The adjusted TFR1 are considerably higher than the observed ones, but they are considerably lower than around 0.9. The latter can be thought of as a minimal level observed under the prevalence of universal childbearing. Ellman (2000) cites2 that 15-20% of couples in Russia are infertile but the support of this information seems insufficient, and the number seems unlikely high. Therefore the end of the 90s is a period of a spread of voluntary childlessness in all countries in Central and Eastern Europe, where the relevant data are available. This is a break of a long-standing and stable reproductive norm. Postponement is a tendency that counteracts the norm of early childbearing. It develops smoother than the abrupt transition to a loss of universality of childbearing and could go on longer. Given an average increase in the mean age of birth of the first child at around 1.5 years for a decade, we may expect that postponement could rise for decades until this mean age reaches levels observed in Western European countries. Rejection of childbearing is much closer to western standards.

Note should be taken that all these inferences are based on the validity of Bongaarts-Feeney formula. Kohler and Philipov (2001) and Kohler (this conference) discuss important extensions that allow for a more precise estimation of tempo effects. Their implementation is not likely to disprove the validity of the general inferences from above but may contribute to a deeper understanding of these important changes in fertility behavior. In addition, note is necessary that the adjusted TFR is a period measure of fertility. The interpretation of the drop in the adjusted TFR1 as a sign of voluntary childlessness derives from expectation that its value is unrecoverable in a cohort perspective; see also the discussion by Frejka and by Lesthaeghe (this conference).

Extra-marital fertility marked a considerable increase during the decade of interest. In most of the countries its initial start from low levels led to unprecedented high levels (table 8). In some of the counrties the rise has begun before the transition, but care should be taken for its interpretation because it could be an artifact of the population policy. In particular, lonely mothers were given some advantages, such as priority in housing supply, tax reduction, etc. The data indicate that towards 1999 the traditional

2 Ellman cites “O nekotorykh mediko-demograficheskikh I sotsial’nykh aspektakh razvitya podrostkov”, Statisticheskoe obozrenie, 1999, N.3.

- 9 - 09/06/02

dominance of marital fertility is gone in many countries from the region. Rychtarikova (1999) notes that a part of the rise of extra-marital fertility in the Czech Republic in relatives terms (number of babies per 100 live births) can be the result of falling marital fertility. The relative rise can be explained by the presence of heterogeneity in the population: a certain sub-population group may traditionally have a high level of extra-marital fertility and have maintained it during the transition period with a relatively low decline in its fertility. Ethnic groups like the Roma-Gypsies that reside mainly in the South-Eastern European countries are a convenient example. This consideration does not explain the whole rise in extra-marital fertility in absolute terms as is evident in table 8.

Another reason for the rise in extra-marital fertility is the inefficiency or non-practicing of contraceptive behavior, particularly among teenagers. Early start of sexual life spread widely during the transition period while usage of modern contraceptives spread at lower rate. Teenage fertility increased tremendously in some countries, to rank them among the highest in Europe. There are no direct data allowing the assessment of this explanation to the rise of extra-marital fertility: statistical information on family planning is still scarce. A third reason for the rise of extra-marital fertility is the conception preceding the marriage; the couple may go into a legal marriage after the birth (see Vikat, 1994 for the high rate of births preceding marriage in Estonia). Finally, there is a part of extra-marital fertility that is to be contributed to extramarital unions, but its share is unknown.

Thus we are not able to assess what pat of extra-marital fertility could be contributed to the appearance of new tastes or break of traditional norms and values. We are left to believe that the above two considerations are insufficient and that there is an observance of break of traditional behavior.

The postponement of marriages and the increase of the share of single persons in reproductive age have their impact on the tempo and quantum of fertility as far as marriage is a prerequisite to fertility. One can see this link the other way round though, like Hoffman-Novotny, in that people do not marry because they do not want to have children. This topic needs additional research in CEE countries.

Finally, one should consider the impact of emigration on the estimation of the fertility rates. During the decade many young people have left their countries, whether for studying or working abroad. The number of emigrants is unknown. In Bulgaria it is considered to be around 700,000 and this is app. 8% of the total population. These people are included in the denominators of the fertility rates. A simulated estimation (just extract 10,000 women from each age group in the age interval 18-28) shows that the observed TFR in 1998 could be 1.3 instead of the registered 1.1.

This paper will not discuss family planning issues. David (1999) provides a discussion related to countries in the region. We will just mark some facts Many countries adopted legal changes related to family planing with the purpose to make it respond to a democratic society. Family planning organizations were initiated and developed. Modern contraceptives became available, although at high price for part of the population. Abortion rates decrease but are still high in comparison to western societies. It is worth noting that the 1993 legalization of abortions had no particular effect on fertility in Poland (Kulczycki, 1995).

The descriptive analysis presented here brings to essential inferences. Fertility trends during the transition differ entirely from those observed during the pre-transition

- 10 - 09/06/02

period. The data support the break of norms that have prevailed during the previous decades: (1) The traditional pattern of early and nearly universal motherhood is replaced by one of a gradually increasing mean age of first birth and loss of universality of maternity.

(2) The two-child family model dissolved as early as during the first few years of the transition. There are no signs of its restoration. (3) Extra-marital fertility increased.

In addition, the description showed that the fertility decline as specified by its tempo and quantum components diverged at the start of the period but converged towards its end. The illustrations provided here make it possible to induce that a short-term quantum shock in the fertility decline was experienced by the countries whose economic adjustment did not rank them among the best ones, i.e. were not in our region (a).

The most striking inference is the speed of the changes. The new trends exploded and spread within a decade. Although the pace differed, towards the end of the decade the fertility trends converged, just like in the end of the 80-s.

3. Basic approaches to the explanations of fertility changes

The sudden change in fertility behavior has been a topic of hot debates in the mass media as well as among scientists from countries in the region. The prevailing explanations can be broadly collected in two groups. One relates to the direct social and economic effect of the transition, while the other emphasizes cultural and particularly ideational changes. We will refer to these two approaches as to the economic one and the cultural one correspondingly.

(a) The economic approach This approach originated from the social and economic hardships experienced by large groups of the population in the countries from the region. The social and economic indicators in tables 1 to 4 and the diverse measures of fertility are correlated, as can be seen without the application of statistical analysis. This visual correlation is viewed as a causal relation, used by politicians and mass media to assess the demographic situation as catastrophic. Philipov and Kohler (2001) provide several examples, an extreme one being the search for impeachment of Russia’s ex-president Yeltzin, accused among other things in population genocide with one item being the fall in the birth rates. Although this view to reasoning the fertility drop is widely shared, it has received surprisingly little scientific support. Scientific publications are country-specific, often descriptive, and discuss the above-mentioned correlation in broader terms. Macura (2000) provides a useful study on the effect of the transition’s economic hardship on fertility in Central and Eastern Europe. Using vital statistics data, he reached a positive conclusion for the existence and significance of such an effect. Cornia and Paniccià (1995, 1996) provide other relevant studies. The economic explanation of the reduction in fertility is based on the escalation in the direct costs of children as a consequence of the collapse in income. It is most relevant to the South-Eastern European and the ex-Soviet Union countries. In Poland, the Czech Republic, Hungary, and particularly in Slovenia, the economic adjustment developed

- 11 - 09/06/02

without any or with low recession and a rise in the costs of children could be observed only for subgroups of the population, for example among the impoverished. Relative deprivation is another social and economic aspect that could have an impact on fertility behavior. Lesthaeghe in several of his papers (1998, for example) provides a discussion in the case of western European countries and in the context of the Easterlin hypothesis. Carlson studied the Easterlin hypothesis in some CEE countries during the last few decades before 1989 and found an inverse relationship between the relative size of cohorts and the level of fertility. His explanation is based on the extensive component of the economic growth in these countries, i.e. the larger the young cohorts that enter the labor market, the higher the economic growth. Towards the beginning of the 70s the relative increase in cohort size was exhausted and so was the extensive factor. The regime could not provide for the development of intensive growth, therefore the countries experienced rising economic problems and hence fertility exhibited a tendency to decrease.

The start of the transition destroyed this relationship, as could be expected with the pertinent economic restructuring. Relative deprivation acquired a different aspect, a cross-sectional one, to distinguish from that based on relative cohorts. Many people did not experience a significant drop in their personal well-being but considered themselves economically deprived relative to others. For example, some people with higher education than others could feel deprived because they received lower income. Another example is provided by the disappointment in people who have expected a more significant improvement of their economic situation with the arrival of a more democratic regime. In general, relative deprivation would mean that people assess their income as low relatively either to others or to their personal expectations, and hence view their economic situation as inconvenient for having a(nother) child. Rychtarikova (personal communication) considers relative deprivation to be an important cause for the fertility decrease in the Czech Republic. Sztompka (1996) discusses relative deprivation “after revolutionary elation and unrealistic hopes” (p.37).

Macura (2000) indicates the importance of increase in education on postponement and level of fertility. The impact of longer education on fertility in Western European countries has been widely studied. Blossfeld and Huinink (1991) suggest a theoretical framework that seems applicable to the CEE countries. The discussion on education in the first section revealed that there is a significant rise in the enrolment ratio and in the length of education among young adults that possibly has taken place in many CEE countries. The higher level of education is likely to invoke an individual’s preference to a job career. Therefore an augmenting circle of women are expected to postpone family events including births. This aspect of the transition could be an important reason for the recent fertility drop. Given the structural change on the labor market and the increased demand of highly qualified professionals it can be expected that educational expansion and its catalytic effect on job commitment will keep fertility low for quite some time in the future, at least where young adults are considered.

Rising economic uncertainty is another aspect of the economic approach, whose relevance to the fall in fertility in Eastern Europe has been theoretically discussed by Ranjan (1999). The discussion is particularly important because it applies to the countries that did not experience a profound economic recession. Nauck and Joos (1995) indicate

- 12 - 09/06/02

the primary importance of “unstable expectations for the future and shifts in life orientation” for the explanation of the abrupt fall of fertility in the former GDR. Lack of housing has been a traditional reason for a preference to a low family size in the CEE countries. Its significance has increased during the recent decade, due to the lower income of people who therefore cannot afford buying a dwelling. Frazscak (2000, p.32) indicates the need of housing and its impact on fertility in Poland.

Deteriorating economic situation, education expansion, expensive housing, and rising economic uncertainty were discussed here as causes for the drop in fertility. It is likely that at the micro level couples placed under the pressure of any of these factors will take a decision to postpone crucial life events, such as a marriage and a birth. Hence the trends discussed in the previous chapter may all be relevant to the economic approach.

(b) The cultural approach This approach refers primarily to ideational changes, that is, changes in norms, values, attitudes that cause consequential behavioral modifications. Other aspects of culture, such as those that refer to habits and traditions, may have an impact on fertility change but will not be considered here because of the lack of information and relevant research. Some authors discussed the “cultural decommunization”, although the link to fertility is not immediately clear. Indeed, folklore and folk traditions are usually pro-natalistically oriented, and one may suppose that when traditions modify their normative pressure for higher fertility may change. The impact of cultural changes on fertility behavior has been widely studied where Western European and other developed societies are considered (Lesthaeghe 1983, Hammel 1990, and others). The role of cultural factors on falling fertility has been convincingly illustrated. These factors have a long-standing effect and develop for many decades and even centuries. Secularization, rising female autonomy, rising expressive individualism, are among the most relevant ones. They contribute to a lessening pressure of traditional norms pertaining to high fertility. Lesthaeghe and Vad de Kaa (1986) and Van de Kaa (1987) suggested the term “second demographic transition” to denote significant changes in demographic trends that occurred in Western Europe since the middle of the 60s. Where fertility is considered, these trends are alike the ones described in the previous section. The authors emphasize the importance of ideational and particularly value changes for the explanation of these trends. The term gained large popularity among demographers in the CEE countries and particularly in the Central European countries. A recent book entitled “Demographic change in Poland during the 90s in the light of the second demographic transition” provides a vivid example (Kotowska, 1999). The SCR-s (see footnote 1 in the previous section) provide country-specific discussions. Some demographers from the CEE countries address the issue in their work published in the countries’ population journals. The impact of the ideational changes to fertility decline in the CEE countries is discussed usually the way it is discussed for the Western countries. Long-term secularization, rising individualism and female autonomy are seen as trends that have developed during the years of the totalitarian regime. The latter may only have hindered the fall of fertility through population policies. The start of the transition accelerated the ideational changes and the connected fertility decline.

- 13 - 09/06/02

Studies on the second demographic transition in the CEE countries usually link it to the trends exhibited by vital statistics data. Since these trends are similar to those observed earlier in Western European countries, it is a straightforward analogy that the second demographic transition has been as effective in the eastern part of Europe as in its western part, although exhibited with some delay.

The implementation of the “second demographic transition” approach to CEE countries is usually set in the framework of the overall changes in life and society. Thus it comprises more than just pure ideational changes. For example, the discussions evolve around female and young adults’ labor force participation, conflicting roles of a working mother, rising education. As a consequence it interacts with the economic approach discussed above. Such an interaction is not surprising given the commonality between social, economic, and cultural change. Lesthaeghe (1998) discusses this topic from the theoretical viewpoint and his inferences serve as a theoretical justification of a similar approach.

4. Economy or culture?

The availability of two distinct explanatory approaches for the fertility drop raises a number of questions like the following ones. Which one of them is the true one? Do they interact and how? Are the approaches valid for sub-populations or both refer to individuals at the same time?

The search for an answer to similar questions encounters a significant problem, namely the lack of data. Research directed towards verification of the approaches is based on aggregated-level demographic data that describe general trends like those discussed in sections 2 and 3 above. The usage of one and the same data set for the support of two distinct theoretical approaches indicates that data are insufficient for the better study of interactions between social-economic and ideational factors. No country in the region disposes with an appropriate sequence of fertility surveys that could be used for the verification of theories and approaches to fertility. The FFS are the only ones available in the region.3 Since they provide event histories for births, cohabitation, study, and work, it becomes possible to trace the interrelation of events along different cohorts.4 The FFS data have been used by researchers in the countries where they are available for the purpose to enrich the knowledge on demographic behavior and less so for their explanation.

Macro-level data could be useful were they available for population subgroups that are not expected to be subject to the impact of one of the two grand approaches. For example the population in some remote regions is not expected to be subject to ideational change, and such regions are usually the poorest in a state. Unpublished results of the author indicate that the Bulgarian Muslims and the ethnic Turks in Bulgaria experience a significant fertility decline and postponement of births (as well as of marriages). These demographic trends cannot be linked to ideational changes.

The lack of adequate data leaves open issues of application of the two approaches.

3 Some statistical offices have conducted their surveys on fertility. They provide information that is unfortunately less relevant to the discussion here. 4

A comparative study for the countries in the region is initiated by the author and is underway.

- 14 - 09/06/02

Consider the economic approach. It relies mainly on the impact of changes in the costs of children. Indeed, when people struggle for surviving it is likely that they will view the birth of a child as bringing a competitor to the scanty family resources. In real life individuals range from this extreme situation to the reverse extreme one, i.e. sudden enrichment. We know nothing about the level of fertility differentiated by this wide spectrum of economic situations. The international data prompt that the relative number of individuals in a situation of struggling for survival is higher in countries where the economic situation is relatively worse. We would expect therefore a relatively lower fertility level in these countries, and this is not observed.

Consider the cultural approach. Ideational changes are long-standing and have started their development well before the start of the transition period. Then one may ask how could modern value orientations develop in a suppressive regime like the totalitarian one. The regime favored the rise in conformism and alienation, rather than in individualism and autonomy. Some modern values, such as female autonomy and secularization, were indeed widely spread over the region. The reason is in that they were imposed to the people. Female autonomy rose due to the significant increase in the share of working women. This increase was a consequence of the constitutionally guaranteed right of labor and the availability of working places. Hence female (economic) autonomy was a status achieved through the specifics of the regime, rather than through changes in values. The regime imposed restrictions on practicing religion. Although one may suppose that religiousness remained an internal value, unshared to others for the fear of sanctions, norms based on religion gradually weakened. Thus the regime contributed to the development of secularization.

5. The transition period - discontinuity and disorderliness

The previous sections demonstrate that the evidences for the economic or the cultural approaches are sparse. While the lack of adequate data is one important reason for this observation, another one is connected to their argumentation. In this section we discuss aspects of the latter. The discussion here starts from the comprehension of the transition period as one of a social crisis, although with different attitudes.

The first chapter revealed discontinuity as a major characteristic of the transition period. It is manifested through the reconstruction in political, social, and economic structural organizations and other institutions from totalitarian to democratic ones. In the context of the CEE societies, discontinuity means that old institutions are demolished and new ones are constructed. The abolition of the old is a momentous event, while the construction of the new is not. The new institutions appear in an environment that is not perfect for their functioning, nor are they constructed free of errors. Economic aspects of discontinuity, termed “hiatus”, were studied by Kozul-Wright and Rayment (1997). Consider for example the legal system. It is impossible for any National Assembly to change many hundreds of laws. This takes place throughout time. New laws are often imperfect and require further amendments. Thus diverse niches in life experience the impact of lawlessness. Such a state of legal anomie has been vividly described by Sztompka (1996, p.55): “The legal system is a fragmented mozaic of partial regulations, old and new, often inconsistent, repeatedly changed, arbitrarily interpreted. …The rule of

- 15 - 09/06/02

law is compromised by extra-legal decisions taken by the highest authorities…including retroactive legislation.. by the parliament. Trust in continuity, stability and the orderliness of social life is effectively undermined.” This description was relevant at the time to most of the countries in the region.

Discontinuity refers to the period of time that starts with the destruction of the totalitarian society and is expected to end with the construction of the democratic one. Discontinuity distinguishes the CEE societies from more developed and less developed ones.

Discontinuity is pronounced in cultural shifts as well. The sudden overthrow of a totalitarian ideology brings about breaks in traditions, norms, and values. From the point of view of ideational shifts it represents a relatively short period of time, if not a moment, when previously existing norms and values have been broken and newer ones have still not been settled. This is a state of deinstitutionalization and normative deregulation, known as social anomie (normlessness), following Durkheim’s tradition in sociology. It has been the subject of research among social scientists in CEE countries. Genov (1998) studied the impact of anomie on quality of life in Bulgaria. Using survey results he found high levels of reported uncertainty, anxiety, and fears. He classified the latter as telling an anomic situation. Rousselet (1994) discussed the lack of identity of the Russian society in the context of anomie, and the role religion could play in the search for identity. Arts, Hermkens, and Van Wijck (1995) viewed disorderliness in the CEE countries in the light of the theory of anomie. The Czech journal “Sociologický časopis” devoted one of its issues to the topic of transformation as a social anomie (Burianek, 1994). RabuÅ¡ic and MareÅ¡ (1996) used Srole’s (1956) index of anomie to find from two independent surveys that while in November 1994, 29% of the studied population was anomic, in June 1995 this share rose to 36%. He found also that anomie increased with age and decreased with education. His conclusion is that the longer the time period since the “velvet” revolution of 1989, the less enthusiasm the people have with respect to the transformation of society. Sztompka (1996) viewed social disruption as the “deficiency of cultural resources, the central of which is trust”. “Widespread anomie” and “relative deprivation” have brought about the spread of a “culture of distrust”.

Arts, Hermkens and Van Wijck (1995) base their analysis on the Durkheimian theoretical framework of anomie. For the purposes of this paper it is necessary to cite some of the founding views in this framework. In their words, “anomie… is a particular form of deinstitutionalization, … a sudden normative deregulation of society… …a particular form of individual disorientation. In a situation of anomie people are thrown back upon their own elementary behavioral propensities. This is due to the absence of a stable, external frame of reference which can restrain their needs and desires.” (p. 2). Individual disorientation arises when “individual attitudes and actions have been disconnected from shared values, norms and behavioral standards” (p.3). The breakdown of societal rules brings about social disruption. “Uncertainty will increase because people no longer know what is still feasible or not, or what is or is not just, or what demands and desires cannot be justified.” (p.3).

Discontinuity, anomie, or disorderliness are not specific to the CEE countries only. They have been a subject of discussion in the broader context of development (see for example Goulet, 1992).

- 16 - 09/06/02

Disorderliness has an impact on fertility through several aspects, an important one being the high level of uncertainty felt by the individuals. Placed under the pressure of uncertainty, people postpone irreversible events such as births. Some individuals will postpone till their full rejection. Uncertainty thus causes postponement of births and fall in fertility. Uncertainty effects in a similar way marriage: people may prefer the less risky state of cohabitation, given the normative and legal commitments to marriage. Extra-marital fertility may rise in such extra-marital unions.

Discontinuity and disorderliness need to be considered in the study of the economic and particularly the ideational approach to fertility change in CEE countries. In such a state people may pertain to the surviving norms that are usually traditional ones. Inversely, disorientation may increase the variance of pertaining to a particular common value or norm. Such value changes need to be separated from the long-standing ideational change towards a modern society.

Discontinuity and disorderliness need not encompass all the people in a population, nor is their impact equally strong. Rabušic and Mareš (1996) discusses eunomie and semi-anomie as denoting weaker states of anomie. As mentioned above, he identified different groups of the population according to the level of anomie.

People react differently to worsened economic situation, break of norms and values, and uncertainty in general. Some of them remain passive, others take up specific actions. The latter are known as coping strategies. Examples of widely spread coping strategies are finding additional job (second and often a third work), emigration, increase in education, entrepreneurship. Other examples of coping strategies are theft, tax evasion, corruption and bribe. An example of an economically passive coping strategy is working on own plot of land. The EBRD (2000) “Transition Report”, table 5.3, p.103, finds a high significance of subsistence informal work among ex-USSR countries and some countries situated in South-Eastern Europe.

Individuals that develop successful coping strategies ease the economic pressure on themselves and their family. They may also feel less uncertain about the future. Therefore successful coping strategies may weaken the impact of uncertainty and low income on fertility.

Philipov and Shkolnikov (2001) went further in their study of fertility intentions in Russia. Using data from the Russian Longitudinal Monitoring Survey, they found successful those coping strategies that were linked to friends and relatives. Getting help from friends and relatives and providing help to them came to be among the significant factors associated with fertility intentions. Such reciprocal relations refer to the amount of social capital that an individual disposes with. Only this component of social capital was found effective, because others, related to community and diverse state and civic institutions, have suffered the impact of disorderliness.

We need to note another important impact of anomie on fertility. It comes through loss of the desire to have a child. Srole’s measurement of anomie rests on a scale built on several survey questions, one of which refers to “It is hardly fair to bring children into this world with the way things look for the future.” This paper will not consider the complicated topic of measuring anomie; it is just indicative from Srole’s scale that anomic people may have less desire for children.

- 17 - 09/06/02

6. Towards empirical evidence

The discussion in the previous sections implies that fertility change can be causally linked to the following group of factors: (a) economic and social, including measures of increase in costs of children, relative deprivation, rise in education.

(b) ideational, including significant changes in norms, values, and attitudes that have an impact on fertility intentions. (c) related to anomie and disorderliness, causing decrease in the desire to have a child through rise in distrust and uncertainty in the future.

There exists no set of data even for a single country from the region that allows for the composite study of these three groups of factors. An exception is Russia because of the availability of the RLMS. The FFS and the European Value Surveys (EVS) provide indirect and partial evidence. The FFS include a few questions on values but provide virtually no information on the economic status and uncertainty. The EVS include lots of questions referring to value change, economic situation, and uncertainty, but they provide virtually no information on fertility.

Giulia et al. (1999) have done a comparative study for four FFS countries, including Hungary, where the first two groups of factors were considered, explicitly outlining the restrictions of the data set. They find that rise in education and pursuing a career are of particular importance in the four countries; Hungary is exceptional in some respects, for example with its less pronounced religiosity factor.

The tables in the FFS SCR-s (footnote above) reveal additional information on value changes and associated behavior. The age at first intercourse is a variable related to fertility that shows a uniform change among the 8 countries: it decreased markedly with each younger cohort. Sexual life started earlier with each cohort while first union is usually postponed. Such a change in sexual behavior reveals a weakening of a norm that related start of sexual life to a close amorous relationship preceding an eventual marriage.

Van de Kaa (1998) has studied changes in value orientation in a number of countries, including CEE countries, on the basis of the 1990 round of the EVS. He finds convenient evidence in the spread of post-modern values and its association to fertility preferences. Lesthaeghe and Moors (1996) used the EVS to study the impact of value change on living arrangements in several countries from Western Europe.

The EVS provide data from two waves: 1990 and 1995-97. For this reason they are considered in some more detail. We consider the population aged 20 to 29, which are ages of high fertility for the countries in the region of study. We consider males and females together because the sample size is very small. Thus we lose sex-specific differentials in value orientations.

The only information the EVS contain with respect to fertility are one question on the ideal number of children, and one on the number of children the respondent ever had. The latter question refers to events that have happened at time when the respondent might have different values than those observed at the time of the survey. Hence it is inconvenient for inclusion in the study. The ideal number of children is not related directly to fertility behavior, but being a value it can be related to norms, values, and attitudes as revealed by the respondents at the time of survey.

- 18 - 09/06/02

The ideal number of children, as observed in the CEE countries where data are available for both waves, is given in table 9. No statistically significant drop is marked in the countries from Central Europe and Estonia. In the other four countries it is significant: it is to be noted that in Russia this drop is with 0.44, and in Bulgaria it went down below 2.0: the only country on the continent with this low level of ideal number of children, besides the ex-GDR.

The small number of observations forces an aggregation of the countries for the needs of a further analysis, even only a descriptive one. We use the aggregation given in the end of the introduction, with the following changes: further aggregating the South-Eastern European and the CIS countries into one group, the exclusion of Belarus for the lack of data, and the inclusion of the ex-GDR in the group of the Central European countries. Table 9 indicates for which of these countries there exist observations in the 1990 only, in the 1995-97 only, and in both waves. We will denote the group of the Central European countries by “CEC”, the Baltic countries by “BC”, and the third group in South-Eastern and Eastern Europe by “SEE”. Note that each group comprises different countries for different waves. This is a clear deficiency of the study. Weights were designed by the organizers of the survey such that to equalize the sample sizes among the countries.

Consider the ideal number of children in the three groups. In group CEC it did not change significantly from 1990 to 1995, although the number of persons stating one child as the ideal number decreased from 9.8% to 6.8%, the base being correspondingly 906 and 486. In the SEE group this per cent rose from 7.0 to 10.5 from a base of 858 and 1253 correspondingly. In the BC group the per cent rose from 3.7 to 4.7, from a base 589 and 581 correspondingly. Notably, in the 1995-97 wave there appeared 15 persons in the SEE group who stated 0 children as an ideal number (1.7%); 10 persons in the CEC, all from the ex-GDR, and 2 persons in the BC.

Table 10 gives the per cent of respondents who agree with the statement made in questions that are indirectly related to fertility decisions. In all cases, except for three, the changes are towards a widening of the domain of modern values relative to the traditional ones. The results in the table support therefore the ideational change approach. Two of the exception refer to the SEE group, and describe no change with respect to the choice of “agree”. In the case of question 96, there is a significant decrease in the number of respondents who disapprove of the statement. It thus supports the general ideational change.

The “Inglehard” questions provide an interesting counter-example (table 11). In the CEC group there is a slight increase in post-materialism on account of the transitional state (the chi-square p-value is 0.109). In the other two regions there is a return to materialism during the period 1990-1995. This latter case can most probably be connected to the worsening of the economic situation in the states during these five years.5

5 Another explanation is also possible. The Inglehard questions rely on the concept of freedom. The latter, following Locke’s tradition, has two meanings. One is the freedom achieved as a status of an individual. The other is freedom in the liberal sense, one that is restricted by the rights of others. In 1990 all CEE countries experience the freedom from the totalitarian regime and therefore the fist meaning of the word was actual. 5 years later the second meaning may have become actual. Thus the question could have an ambiguous meaning.

- 19 - 09/06/02

Four questions relate to the economic situation of the respondents. The first one gives information on unemployment and study. Reported unemployment is considerably higher than the official estimates presented in table 2. The fact that the data in table 12 refer to persons 20-29 years old does not seem sufficient explanation. It is more likely that persons have reported themselves as unemployed although they might not have been registered as such. The data on the status of “studying” at the time of interview support the macro-level information presented above, where the SEE and the BC groups are considered. In the SEE group of countries the respondents do not report rise in education. In bulgaria and Russia, for example, the only two countries with data in both 1990 and 1995-97, the share of respondents that studied at time of interview decreased from 10 to 7 per cent. The second question refers to saving money (4 categories: managed to save money; just get by; spend some savings; borrowed money). It was asked in the 1995-97 wave only. The results indicate a considerably poorer situation in the SEE group and a better-off situation in the CEC group. The next two questions on the economic status refer correspondingly to self-assessment of income and satisfaction with the financial situation, both on a 10-level scale. The main inference they allow is that the variance around the mean of the answers has increased considerably in the SEE and BC groups, but did not change in the CEE countries.

The EVS contains questions that may be related to anomie and disorderliness. Their usage requires the availability of research on operationalization and measurement that we do not dispose with. Anomie is expected to refer to a certain proportion of each population, whose identification requires additional research.6 In addition to these problems, most of the questions relevant to this topic were asked only in the 1995-97 wave.

We can make a crude assessment of disorderliness by referring to the attitude of the respondents to the political system. A question on the comparison of the political system today (1995-97) and 10 years ago favors the today’s system, but the preferences to the totalitarian regime are significant: some 26% in CEC, 35% in SEE and 23% in BC give a preference to the old regime. The next question asks on the assessment of the political system as it is today. It is 38% of the respondents in CEC, 24% in SEE and 28% in BC that give a positive or a non-negative assessment (points 6 to 10 on a 10-point scale). Finally, the assessment about the future is abundant with “DK”: 15% in CEC, 29% in SE, and 24% in BC. These results indicate a wide variety of views on the political system. Another question refers to the satisfaction with the work of the government. It reveals that only 27% in CEC, 15% in SEE, and 16% in BC would see it as “very” or “fairly” satisfactory, versus fairly or very dissatisfactory.

Several questions refer to the trust the respondent has in diverse institutions: legal system, government, Parliament, among others. Data are available for both 1990 and 1995-97 for some of the questions. The descriptive data indicate in general a change

6 For the purposes of an illustration are given the questions used for the derivation of Srole’s scale: (1) There is little use writing to public officials because often they aren’t really interested in the problems of everyday man. (2) Nowadays a person has to live pretty much for today and let tomorrow take care of itself. (3) In spite of what some people say, the lot of the average man is getting worse, not better. (4) It is hardly fair to bring children into this world with the way things look for the future. (5) These days a person doesn’t really know whom he can count on. (Srole 1956: 712-713)

- 20 - 09/06/02

towards increase in trust where BC group is considered; decrease in trust in the SEE countries, and no significant change in the CEC countries. Trust in governing bodies is indirectly assessed through two additional questions. One is: “..would you say that this country is run by a few big interests looking out for themselves, or that it is run for the benefit of all the people?” There is an enormous rise of respondents answering “run by a few big interests” from 64% in 1990 to 80% in 1995-97 in the CEC group, and from 45% to 91% in the BC countries; in the SEE group it was high, around 75%, in both waves.

The other question refers to the widespread of bribe and corruption (asked in the third wave only). The possible answers are: (1) almost none (2) few (3) most (4) almost all public officials are engaged in it. 43% of the respondents in the CEC countries chose “none” or “few”; 11% in the SEE and 15% in the BC countries.

The descriptive discussion above indicates the following inferences with respect to the three groups of factors mentioned in the beginning of the section. (a) The economic situation is significantly better in the CEC group; there is considerable less indication of its deterioration there as compared to the BC and especially to the SEE countries.

(b) Values related to fertility behavior have changed towards modern orientation in all the countries from the region. (c) There is a significant increase in distrust and dissatisfaction with the political system. There is an indication of increased disorderliness, mainly in BC and especially in SEE group.

The next step in the search for empirical support with the EVS data is the use of a multinomial logit model for the study of the association between the ideal number of children and different variables. The choice if this variable as related to fertility is forced: there is no other convenient variable in the data set. There is only one, rather feeble justification for its usage here: it could be related to fertility behavior, insofar as it decreased just as fertility did, although it is considerably higher.

The description of the observations on the ideal number of children above and particularly the appearance of choices of 0 children and the relative increase in the choice of 1 child, indicate specifics of the transition period, such as disorderliness. We further refer to an ideal number of children equal to 1, 2, or 3 and thus avoid choices that are either extreme or less relevant to the discussion here (for example choices specific by ethnic group). The multinomial logit provides the possibility to trace the impact of different sets of variables on the choice of 1 or 2 children relative to 3. The results, presented in terms of relative ratios, are given in table 13.

The number of children the respondent has ever had was also used as a control variable, not included in the table. This variable could not be used for the CEC group of countries, where the number of children was presumably coded wrongly (no respondents with 0 children ever born).

The variables included in the model were chosen by a simple criterion: represent the three groups of factors described in the beginning of his section. Their choice was done after experiments with other variables that related to each one of the factor groups. Unfortunately principal components or factor analysis did not give convenient solutions

- 21 - 09/06/02

and thus could not help for the search of a latent variable that would describe a group of variables. The first three variables have a traditional significance in the study of fertility. Religiousness and size of settlement represent traditional values. They both are found to have a similar effect on the choice of an ideal number of children, where they are statistically significant. For example, the larger the size of settlement where the respondent lives, the more likely he/she is to chose 1 or 2 children as an ideal number, relative to 3. The drop in religiosity corresponds to a choice of a lower ideal number of children. Both variables are insignificant in the Baltic countries.

Education is significant in all cases, except for the choice of 1 relative to 3 children in SEEC. Increase in education relates to a less preference of 1 or 2 ideal number of children relative to 3. This observation contradicts the association where fertility is considered: higher education usually relates to a smaller number of children. Since we relate here values, and not behavior, possibly higher education means less feeling of insecurity and hence a preference to a higher ideal number of children.

The next two variables, importance of family and friends, are significant in a specific way in the SEEC group only. The less important is family in life, the more probable is the choice of 1 versus 3 ideal number of children. The less important are friends in life, the less likely is the choice of 2 versus 3 ideal number of children. These variables point to the significance which family and friends may have in the formation of fertility preferences. Family and friends relationships are traditionally strong in these countries; their presence enriches the individual’s social capital. Philipov and Shkolnikov (2001) reached a similar finding, where the fertility intentions in Russia were studied with the RLMS data.

The self-assessment of the financial situation of the household reveals an association as could be expected: the better the financial situation, the less likely is the choice of 1 or 2 children versus 3.

The variable on agree versus disagree whether a woman needs child in order to be fulfilled associates very strongly with the dependent variable: those that disagree with the statement, relative to those that agree, are much more likely to point a lower ideal number of children than 3. The choice of 2 versus 3 ideal number of children in BC is an exception that cannot be immediately explained.

The higher the bribe and corruption, the more likely are the respondents in the BC group to prefer 1 or 2 children as an ideal number, relative to 3. This statistical inference is in line with the discussion on disorderliness and uncertainty. Now the SEE countries have an exceptional link in the inverse direction.

Conclusion

The abrupt fertility fall in the CEE countries is a result of the composite effect of numerous economic and ideational factors. The data are insufficient for the delineation of their effect, neither for whole population, nor for sub-groups of the population. For this reason the question posed in the title remains without a precise answer. The scarce data allow international comparisons. The latter reveal that the economic factors are more pronounced in the countries in South-Eastern and Eastern Europe, than in the Central

- 22 - 09/06/02

European countries, while ideational shifts take place all over the eastern part of Europe. It remains unclear how ideational shifts have spread among populations that have been until recently separated from the western democratic societies. Behavioral diffusion needs more time than just a decade; therefore the origins of the ideational shifts are likely to be found within the societies in transition.

Both the economic and the cultural approach will benefit by considering primary characteristics of the transition, such as discontinuity, disorderliness, and anomie. They give rise to uncertainty, which leads to the postponement and reduction of fertility. People that manage to cope with this situation will be less uncertain in their future, although their income may remain low. Research on the implementation of the economic approach may focus more deeply on coping strategies rather than on income and work alone. Disorderliness breaks norms and values, and thus may create conditions for a sudden ideational change, aside from diffusion or long-term development. The latter two hypotheses result from the discussion in this paper. Their check requires data that are not unavailable.

The transition has diverse paths in the CEE countries. These countries tend to be more and more divergent, having started from very close positions before the end of the totalitarian regimes. On the other hand, the fertility trends converged. Thus a demographic convergence is the result of social, economic, and political divergence.

Imagine on map 1 a line that unites the cities Sankt-Petersburg and Dubrovnik, instead of Trieste. To the west of this line, and around Hajnal’s line lie the countries who are doing better with respect to the transition. These are the countries from the Central European region and from the Baltic region. One can see to the east of the line well delineated the countries from South-Eastern Europe and the Eastern European CIS countries. We have worked with these regions and have found similarities and differences among them. In the conclusion it just suffices to state that the countries to the east of the Sankt-Petersburg-Dubrovnik line share a common prevailing religion, namely Christian Orthodox. Possibly such a cultural differentiation could contribute significantly to the further understanding of the complexity in the interactions among economic and cultural factors on one side, and fertility decline, on the other.

- 23 - 09/06/02

References:

Arts W., P. Hermkens, P. Van Wijck (1995). Anomie, distributive justice and dissatisfaction with material well-being in Eastern Europe. IJCS 36,1-2: 1-16. Blossfeld, H.-P. and Huinink, J. (1991). Human Capital Investments or Norms of Role Transition? How Women's Schooling and Career Affect the Process of Family

Formation, American Journal of Sociology, 97/1, pp.143-168. Bongaarts, J. and Feeney, G. (1998). On the Quantum and Tempo of Fertility. Population and Development Review 24:271-292. Cornia, G. and R. Paniccià (1995). The demographic impact of sudden

impoverishment:Eastern Europe during the 1989-94 transition. UNICEFF, Florence, Innocenti Occasional Papers EPS 49. Cornia, G. and R. Paniccià (1995). The transition’s population crisis: An econometric investigation of nuptiality, fertility and mortality in severely distressed economies.

Most, N.1.

Council of Europe (2000): Recent Demographic Developments in Europe. Strasbourg. David, H. (1999). From Abortion to Contraception. Greenwood Press. Di Giulia P., R. Lesthaeghe, G. Moors, A. Pinelli (1999). Fertility tempo and quantum: an empirical test of major theories with data from four FFS-countries, IPD-Working

Paper 2.

EBRD (2000) “Transition Report”. European Bank for Reconstruction and Development. Ellman M. (2000) The social costs and consequences of the transformation process. Papers from the ECE Spring Seminar, 2000. ECE, United Nations, Geneva. Frątczak E. (2000). Changes in fertility pattern in Poland during the transformation

period. Polish Population Review, 16. Genov, N. (1998). Transformation and anomie: problems of quality of life in Bulgaria. Social Indicators Research 43: 197-209. Goulet, D. (1992). Development: creator and destroyer of values, World Development,

Vol. 20, issue 3, p.467-475. Hammel, E. (1990) A theory of culture for demography, Population and Development Review 16/3: 455-485. Kohler, H.-P. and D. Philipov (2001). Variance effects in the Bongaarts-Feeney formula.

Demography, 38/1, 1-16. Kotowska, E. (2000). Demographic change in Poland during the 90s in the light of the second demographic transition (in Polish). Szkola Glowna Handlowa w Warszawie. Kozul-Wright, R. and P. Rayment (1997). “The institutional hiatus in economies in

transition and its policy consequences”, Cambridge Journal of Economics, Vol. 21, N. 5, pp. 641-661. Kučera, T., Kučerová, O. Opara, E. Schaich (2000). New demographic faces of Europe. Springer. Kulczycki A. (1995). Abortion Policy in Post-communist Poland. Population and

Development Review, 21(3). Lesthaeghe, R. (1983). A century of demographic and cultural change in Western Europe: an exploration of underlying dimensions. Population and Development Review, 9/3, 411-434.

- 24 - 09/06/02

Lesthaeghe, R. (1998). On theory development and applications to the study of family formation. Population and development review, 24/1:1-14. Lesthaeghe R. and D. van de Kaa (1986). “Twee demografische transities”, in Lesthaeghe and van de Kaa (eds.) Bevolking, groei en krimp. Deventer: Van Loghum Salerus,

19-68.

Lesthaeghe, R. and G. Moors (1996). Living arrangements, socio-economic position, and values among young adults: a pattern description for France, West Germany, Belgium, and the Netherlands, in D. Coleman (ed.): Europe’s population in the 1990s. Oxford University Press. Macura, M. (1999): Fertility decline in the transition economies, 1982-1997: political,

economic and social factors. Chapter 4 in: Economic Survey in Europe, 1999/1, pp. 181-194, UN ECE, Geneva. Macura, M. (2000): Fertility decline in the transition economies, 1989-1998: economic and social factors revisited. Chapter 6 in: Economic Survey in Europe, 2000/1, pp.

189-207, UN ECE, Geneva. Milanovic B. (1998). Income, inequality and poverty during the transition from planned to market economy, World Bank, Washington D.C. Monnier, A. (1996). The demographic impact of the transition in Central and Eastern

European countries. Courrier des Pays de l'Est, 409, pp.74-91. Monnier, . and J. Rychtaríková (1992). The Division of Europa into East and West. Population: An English Selection. 4: 129-160. Nauck, B. and Joos, M. (1995). East Joins West: Child Welfare and Market Reforms in

the Special Case of the Former GDR. UNICEFF, Florence, Innocenti Occasional Papers EPS 48. Philipov, Dimiter and Hans-Peter Kohler (2001). Tempo effects in the fertility decline in Eastern Europe: evidence from Bulgaria, the Czech Republic, Hungary, Poland,

and Russia, European Journal of Population, forthcoming. Philipov, D. and V. Shkolnikov (2001). Fertility intentions and social capital-based coping strategies: results from the Russian longitudinal monitoring survey. Paper to be presented on the PAA meeting 2001, Washington D.C. Ranjan, Priya (1999). Fertility Behaviour under Income Uncertainty, European Jouranl

of Population, 15, pp.25-43. Rabušic, L. and P. Mareš (1996). Is Czech Society Anomic?, Sociologický časopis, XXXII:2, 175-187 (in Czech). Rousselet K. (1994). Anomie, recherche identitaire et religion en Russie. Social compass

41/1, 137-150. Rychtaríková, J. (1999). Is Eastern Europe experiencing a Second Demographic Transition? Report presented at the Lowest-Low Fertility Workshop, Max-Planck Institute for Demographic Research, Rostock, Dec. 1999. Sztompka P. (1996). Trust and emerging democracy. International sociology, 11(1):37-

62.

van de Kaa (1987). Europe’s second demographic transition. Population bulletin 42,1. Van de Kaa, D. (1998). Postmodern fertility preferences: from changing value orientation to new behavior. Séminaire “Démodinamiques”, INED. Vikat, Andres (1994). Family Formation in Estonia. Finnish Demographic Society,

Helsinki.