Selected Research Papers in Social Change, Education, Labour Market, and Criminology Volume II

Figure 2: Average age of mother at the first child and total births: 1994–2008, Republic of Macedonia

28

27

26

25

24

23

22

21

1994 1996 1998 2000 2002 2008 Yearat first child total birth

s

Fertility in the Republic of Macedonia 20Average age of mothe

r

As can be seen, decreasing of fertility in Macedonia persists among all groups, but the differences between groups are related to the timing of the decrease.

Many point out the rising costs of raising children – not just higher costs of living, but also the personal costs to parents of deferred professional advancement and individual fulfillment. Another explanation focuses on related changes in social expectations about marriage and family formation: young Macedonians may not feel the same social pressure to get married and have children as their parents and grandparents did.

4.2 Ethnicity and Fertility

To depict fertility differentials by ethnicity4, cumulative fertility will be used, based on the Census data from 2002. The mean numbers of children ever born to the number of women, obtained from the Census of 2002, are shown in Table 1.

A woman by the end of her reproductive period at the time of the Census (2008) had an average completed fertility of 2.31, which is lower than in 1994, when the completed fertility was 2.5 children.

Table 1: Average number of children ever born, 2002, Republic of Macedonia Age group of mother

15–19 20–24 25–29 30–34 35–39 40–44 45–49 Republic of Macedonia 0.04 0.40 1.19 1.83 2.14 2.26 2.31

Ethnicity of woman Macedonian 0.03 0.36 1.05 1.59 1.84 1.93 1.97 Albanian 0.02 0.37 1.35 2.20 2.74 3.09 3.40

Turkish 0.09 0.69 1.60 2.21 2.54 2.80 3.01 Roma 1.13 1.19 2.06 2.69 2.88 3.08 3.45

Vlach 0.00 0.13 0.65 1.34 1.69 1.75 1.73 Serbian 0.03 0.33 0.99 1.55 1.80 1.91 2.01 Boschnajk 0.03 0.44 1.10 1.72 2.32 2.56 3.09 Other 0.06 0.52 1.24 1.75 2.04 2.14 2.20

Source: Calculated by the researcher based on the 2002 Census data

Ethnicity is the most distinctive feature of fertility in Macedonia and any discussion of fertility would be incomplete without a brief examination between ethnicity and fertility. The initial assumption is that the number of children ever born is bigger among non-Macedonian women. As can be seen, all ethnic groups show a smaller

4 Macedonia is multiethnic country. Out of a total number of 2,022,547 inhabitants, 64.2% are Macedonian, 25.2% Albanian, 3.9% Turks, 2.7% Roma, 1.8% Serbs, 0.1% Vlachs, 0.9% Bosniaks, and 1.1% Other. These ethnicities have different languages and religions.

Fertility in the Republic of Macedonia 2

1

number of CEB for each age group. More precisely, the Macedonian woman, as part of the majority of the population, finished her fertility with an average of 1.97 children in 2002. Less than that number is the Vlach woman with 1.73 children. Little above the number of children for the Macedonian woman is the Serbian woman, with 2.01 children. The Roma population is characterized by the highest fertility. The average Roma woman completed her fertility with 3.45 children and the Albanian woman 3.40 children. Somewhere in between are the Turkish woman with 3.01 children and the category “other” with 2.58 children (see Table 1).

The comparison of data from the aspect of different age groups shows that differences and oscillations are not very significant among women at the younger ages, from 15 to 24 year old, with the exception of the Roma woman. The magnitude becomes wider as the age increases (see Table 1).

The comparison of data on the average number of children ever born at the end of the reproductive age for the whole country (2.31) indicates different dynamics among different ethnicities to reach this number. The Macedonian woman does not reach this number during her reproductive period, and nor does the Vlach or the Serbian woman. On the other hand, the Roma woman reaches this number between the ages of 25 and 39, and the Albanian and the Turkish woman reach the number of 2 children or reach the level of cumulative fertility for the whole country between the age of 30 and 39.

In the case of Macedonia, the explanation seems to be that the observed differences between ethnic groups have a basis on various social, economic, and other characteristics. In addition, certain fertility differences among women of different ethnicities are due to divergent views on marriage and birth control method. Among the group with high fertility rates, a smaller number of women are employed, the level of education is lower, and the family and household also tend to be more patriarchal in structure.

4.3 Education and Fertility

Undoubtedly, education is also a variable with a great influence on fertility. Much has been written about the relationship between education and fertility. In explanations related to differences in fertility, the behavior according to the educational level (Cochrane, 1979) emphasized the fact that education operates effectively at a certain level of social development.

In the Macedonian context, education is found to be an important determinant of fertility in the country and is considered one of the fundamental reasons for the decrease in fertility. In each stage of demographic transition, women with higher education levels had a relatively low fertility rate (Breznik et al., 1980). Such a thesis is persisting.

Table 2 below presents the cumulative fertility in Macedonia by educational status of woman.

Fertility in the Republic of Macedonia 2

2

Table 2: Cumulative fertility by education status of woman, Republic of Macedonia, 2002

5Age of mother Educational status 15–19 20–24 25–29 30–34 35–39 40–44 45–49

Without education 0.37 1.17 1.94 2.43 2.83 3.13 3.51 Uncompleted primary education 0.23 0.98 1.92 2.55 2.94 3.08 3.11

Primary and lower secondary education

Upper secondary education 0.03 0.61 1.49 2.16 2.47 2.54 2.45

0.03 0.25 1.06 1.63 1.87 1.91 1.90

Higher education 0.00 0.18 0.91 1.49 1.79 1.86 1.86 Faculty, Academy 0.00 0.08 0.38 1.04 1.50 1.67 1.70 Master’s degree 0.00 0.09 0.29 0.75 1.32 1.36 1.64 Doctorate 0.00 0.00 0.00 0.48 1.46 1.50 1.55

Source: Calculated by the researcher based on the 2002 Census data

Overall, the completed fertility is higher among women without education and those who did not finish primary education (3.51 and 3.11 children, respectively). These two groups are followed by the group of women with primary education, reported to have 2.45 children. The other categories of women with higher levels of education completed their fertility with less than two children: secondary educated women with 1.90, higher educated with 1.86 children, women with university education 1.70 children, and those with a master’s degree 1.64 and PhD 1.55 children (see Table 2).

The comparison between completed fertility and the educational status of women at the reproductive age provides empirical proof for one of the answers about different reproductive behavior among ethnic groups in the country.

According to the Census data that refers to educational status, the Roma woman has the lowest level of education. Out of the total number of Roma women at the reproductive age, 57.1% do not have primary education, and 36.1% have primary school education. Only 5.6% finished secondary school. The second group with high fertility is Albanian women. The most frequent level of education for these women is primary education. Out of the total number, almost three-quarters of the Albanian women at the reproductive age have only finished primary school, and 13.4% are without or education or have not completed it. Only one-tenth of these women finish secondary education. Just 2.3% complete higher or university education and a master’s degree. More than half of the Turkish women at the reproductive age finish only primary school; Figure 3 shows that 74.0% finish secondary education and 31.0% are without education or have not completed it. Only one-tenth of these women finish secondary education. Only 2.2% complete higher or university education and a master’s degree.

5 Educational status means the type of school where the person has acquired the highest level of education. Present classification follows the classification from the State Statistical Office of the Republic of Macedonia.

Fertility in the Republic of Macedonia 2

3

In contrast, the women who have lower fertility levels have a higher level of education. More than half of Macedonian women complete secondary school. The same experience is exhibited among the Vlach and Serbian women as well. Among them, Vlach women have the biggest share (26.6%) of completed higher or university education and higher degrees, and Serbian women are behind them with 15.8% and Macedonian women with 13.8%.

An important objective of this study is to specify the main determinants of the reproductive behavior, reflected in the number of children ever born, and also the direct and indirect effects of each of these variables need to be identified. This will be achieved through the application of appropriate statistical analysis techniques such as path analysis, which will be introduced in the next section.

5 Path Analysis of Fertility

5.1 The Mathematical Model

The techniques of path analysis (Wright, 1934; Turkey, 1954; Duncan, 1966; Werts and Lind, 1970; cited by Jiseph et all., 1992) enjoy great popularity among social science and among demographers as well. One of the reasons is that it measures the direct and indirect effects that one variable has upon the other, which is emphasized as an advantage of this technique. One of its disadvantages is that it depends on a prior causal scheme, i.e. it requires a well-defined causal ordering with a well-specified causal modeling of the variables used in the analysis (Kendall and O'Muircheartaigh, 1977).

The intention of the path analysis is to combine the quantitative information given by the correlations with such qualitative information that may be handed on to causal relations to give a quantitative interpretation (Dillone and Goldstein, 1984). Two methods are known for estimating path coefficient. The one is the decomposition of observed zero-order correlation between the variables in the system, and the other is the application of the Ordinary Least Square (OLS).

5.2 Measurement of the Variables, Results, and Analysis

Table 3 presents the definitions of the seven independent variables, selected by regression applying the step-wise method, the causal order, and the method of measuring each of them.

The first three variables (age of woman, place of residence, and ethnic belonging) are considered exogenous variables. This means that they are not affected, but affect some other intermediate variables.

Fertility in the Republic of Macedonia 2

4

Table 3: Variables name, abbreviations, causal order, and measurements of the variables in the path analysis, RM, 2002

Variable Names Abbreviation Causal Order Measurement Age of Woman AW 1 Single Years

Place of Residence PR 2 Dichotomous 0 = Rural 1 = Urban

Women’s Ethnicity WE 3 Dichotomous 0 = Other 1 = Macedonian Women’s Education Wed 4 Number of Years of Schooling Age at First Marriage AFM 5 Single Years

Desire Children DC 6 Discrete Number of Children Contraceptive Use CU 7 Dichotomous 0 = No 1 = Yes

Children Ever Born CEB 8 Discrete Number of Children

Table 4 presents some of the features of the distribution of sample by categories of variables used in the path model. From Table 4 it can be observed that 64.4% of the women in the sample lived in urban centers, and the rest (35.6%) in rural areas. In terms of ethnicity, 65.3% of the study population are Macedonian, 25.1% are Albanian, and 9.7% other. Macedonian women are predominantly found in the urban areas (72%) and the majority of Albanians in rural areas (34%).

This percent distribution structurally corresponds largely to the structure of the population by place of residence and ethnic origin. The same table reveals that the median age of women in this survey is 32 years.

Fertility in the Republic of Macedonia 2

5

Table 4: The percent distribution of the variables used in the path model in the study population, Republic of Macedonia, 2002

Place of Residence Rural Urban Total Characteristics

(%) N (%) N (%) N Overall 35.6 342 64.4 620 100 962

Age of Women 15-19 6.7 23 3.2 20 4.5 43 20-24 18.4 63 16.1 100 16.9 163 25-29 19.9 68 19.5 121 19.6 189 30-34 16.7 57 20.2 125 18.9 182 35-39 18.1 62 17.7 110 17.9 172

40-44 11.1 38 13.7 85 12.8 123 45-49 9.1 31 9.5 59 9.4 90

Median 32 33 32 Age at First Marriage

<17 20.8 71 6.0 37 11.2 108 18-19 19.0 65 16.1 100 17.2 165 20-24 55.8 191 54.7 339 55.1 530

25+ 4.4 15 23.2 144 16.5 159 Median 20 22 21

Contraceptive Use Not Using 53.2 182 44.8 278 47.8 460

Using 46.8 160 55.2 342 52.2 502 Women’s Education

No education 6.1 21 3.4 21 4.4 42 Primary 44.7 153 17.1 106 26.9 259

Secondary 42.1 144 51.9 322 48.4 466 University+ 7.0 24 27.6 171 20.3 195

Women’s Ethnicity Macedonian 52.9 181 72.1 447 65.3 628

Albanian 33.6 115 20.3 126 25.1 241 Other 13.5 46 7.6 47 9.7 93

Source: Computed by the research using Republic of Macedonia Survey: 2002

A median AFM of woman is 21 years, and it is explicitly higher by two years among urban women (22) than among rural women (20). The large majority of women (83.3%; 95.6% in rural and 76.8% in urban areas) got married before the age of 25.

Table 4 also reveals that 52.2% of the respondents were currently contraceptive users at the time of survey. Among urban women, this figure was 55.2%, compared to 46.8% among rural women.

Fertility in the Republic of Macedonia 2

6

In terms of education, 4.4% of women did not finish any level of education (6.1% rural and 3.4% urban), and 26.9% finished primary education, whereby this proportion was higher among rural women (44.7%) than among urban women (17.1%). Secondary level of education was completed by 48.4% of women (42.1% in rural and 51.9% in urban areas). University degrees were completed by 20.3% of women, with an explicitly higher proportion of women in urban (27.6%) than in rural (7.0%) areas.

The equations for this model are going to follow the causal ordering, and as a recursive, regression will be fit separately. The regression equitation is as follows:

1. WEd = P41·AW + P42·PR + P43·WE + u. 2. AFM = P51·AW + P52·PR + P531·WE + P54·WEd + u. 3. DC = P61·AW + P62·PR + P63·WE + P64·WEd + P265·AFM + u 4. CU = P71·AW + P72·PR + P73·WE + P74·WEd + P753·AFM + P·DC + u4. 5. CEB = P81·AW + P82·PR + P83·WE + P84·WEd + P8576·AFM + P DC + P87·CU + u5. 86

The Pij denotes the standardized path coefficient of the effect of the independent variable. The ij denotes the standardized path coefficient of the effect of the independent variable (j) on the dependent variables (i) keeping other variables constant, and ui denotes the residual path coefficient which measures the proportion of the standard deviation of the dependent variables which is caused by unmeasured extraneous variables (Bashir, 1982)

2In the present context, regression analysis or OLS is run one by one and path coefficients are estimated to be the standardized regression coefficients and the residual parameters are estimated to be square root of (1-R), where R2 is the coefficient of determination, which explains the amount of change in the dependent variable due to explanatory variables in each equation.

The direct and calculated indirect as well as total effect of the mentioned independent variables on the response variables are presented in Table 5.

Table 5: Summary of direct, indirect, and total effects of selected variables on children ever born, RM, 2002

Variables Direct Indirect Total AW 0.602 -0.001 0.601 PR -0.102 -0.098 -0.199 WE -0.270 -0.103 -0.374 Wed -0.169 -0.057 -0.226 AFM -0.153 0.030 -0.123 DC -0.104 0.000 -0.104 CCU 0.016 0.025 0.041

Source: Computed by the research using Republic of Macedonia Survey, 2002

Fertility in the Republic of Macedonia 2

7

Among all variables which were entered in the model, the largest direct effect on the CEB is the age of the women. This variable has a relatively strong influence on CEB, estimated at 0.602.

Women’s ethnicity (WE) shows the second largest effect with negative sign (-0.270), which means that Macedonian ethnicity has the effect of a decrease in the level of fertility.

Women’s education (Wed) shows the negative effect of -0.169 on the CEB. As woman’s education increases for one year it has a depressing effect on the number of children.

Place of residence (PR) with -0.102 indicates the negative effect on CEB. Urban place of residence is associated with lower fertility, demonstrating a high and negative effect on CEB.

6It is interesting to mention that current contraceptive (CU) use has a positive direct effect of 0.016 on CEB. The explanation for such a relationship regarding the first variable is that contraceptive users tend to adopt contraception after they had at least one child. Further, as for contraceptive use in Macedonia, the 2002 survey tells us that almost all women answered that they knew at least one method of contraceptive methods. The proportion of those who have ever used any contraceptive method is 68 per cent. At the time of the survey 52 per cent of women reported that they were using some contraceptive method. This figure is considerably lower than in countries at a similar stage of development. The desired number of children (DC) shows the value of -0.104. The indirect effects of the explanatory variables on CEB are also shown in the table. This effect is estimated to be slightly higher for three variables such as women’s ethnicity (WE), place of residence (PR), and women’s education (Wed).

Furthermore, after adding indirect effects to direct effects of each variable, the total effect strengthens the increased importance of all the variables, except for the demographic group of variables such as age of woman and age at first marriage. Coefficients of the age of women drop by only 0.001. Age at first marriage shows a slightly decreasing importance of 0.123.

The age of women keeps the prime role in determining the dependent variable CEB with an estimated coefficient of 0.601. The other variables show greater importance with the following order: women’s ethnicity (-0.374), women’s education (-0.226), and place of residence (-0.199).

The greater influence of the ethnicity of women is due to the greater influence of this factor on the education and age at first marriage, and the negative effect on the desired number of children.

Education with a positive effect on the age at first marriage and contraceptive use, and a negative effect on the desired number of children, is considered a variable of importance on the CEB.

6 According to data published in the “2001 World Population Data Sheet”, the percentage of women aged 15–49 using contraception for instance in Slovenia is 84%, Slovakia 74%, Hungary 73%, Spain 72%, Czech Republic 70%, and Ukraine 68%.

Fertility in the Republic of Macedonia 2

8

Due to the above-presented facts as well as the information taken from some other research, it could be said that the ethnic factor is an important factor in the multi-ethnic society on the decision to have a child. Namely, the influence of the norms and values on fertility behavior is related to the fact that special characteristics of each subculture affect the development of personality in the deceive years of socialization in childhood. Therefore, their influence remains more or less prevalent in adult life, even if the actual economic and social conditions give the family a chance and induce other fertility goals. In addition, we may be able to support the thesis of Lenski (1977) about the power of the cultural norms and values concerning fertility even in the modern society.

Undoubtedly, education is a variable with a great influence on fertility. Much has been written about the relationship between education and fertility. In explanations related to differences in fertility, the behavior according to the educational level (Cochrane, 1979) emphasized the fact that education operates effectively at a certain level of social development. In the Macedonian context, education is found to be an important determinant of fertility in the country and is considered one of the fundamental reasons for the decrease in fertility. In each stage of demographic transition, women with a higher education level had a relatively lower fertility rate. (Breznik et al., 1980)

The rural–urban dimension of residence is prominently associated with differences in human behavior. Among these differences, those pertaining to fertility have been particularly consistent. In all types of societies urban dwellers have a tendency to have fewer children than rural dwellers do. It is no different in the case of Macedonia.

The possible explanations for fertility differences between rural and urban dwellers on empirical as well as theoretical grounds are ranged from differences in personality traits to external living conditions. In this study, an explanation has been attempted in terms of selected proxy variables of fertility, and on the educational level. Namely, urbanization is equal to deruralization, detraditionalization, and a transition from the logic of the group to the logic of the couple. New attitudes, among others birth control, family planning, and the preferences for smaller families, arise in the towns where the soil conditions are more favorable for innovation and it takes a while for such attitudes to penetrate into the village. Therefore, development of the towns, economic emphasis on individual behavior, and urbanization have had and still have a depressing influence on the level of fertility, making place of residence a significant factor affecting fertility and making place of residence a significant factor affecting fertility.

6 Conclusion

Fertility in Macedonia has undergone some important changes. The country’s sharp reduction in the level of fertility was followed by structural changes. Those changes consisted of marrying and bearing children at a later age, in the context of the transformation of the socialist regime towards a country with a market economy, which caused both an economic and a social crisis.

Fertility in the Republic of Macedonia 2

9

Observed differences between ethnic groups have a basis on various social, economic, and other characteristics. Among groups with higher fertility rates a smaller number of women are employed, the level of education is lower, and the family and household also tend to be more patriarchal in structure. At the same time, norms and values related to fertility which are closer to traditional and collective behavior among the non-Macedonian ethnic groups and which belong to the specific subculture characteristics of the ethnic groups should not be overlooked.

In order to examine the factors that are associated with the response variable CEB, the path analysis model was adopted as a methodological approach.

All variables entered into the path model offered a strong and direct effect on the CEB, and the total effect proves an increased importance of all the variables on the CEB, except for the age of women and the age at first marriage, where the total effect of these variables shows a slightly decreased importance. The indirect effect on the dependent variable is found to be slightly higher for three variables, namely the woman’s ethnicity, the place of residence, and the woman’s education. However, the age of the woman keeps the primary role in determining the CEB, with an estimated coefficient of 0.601. From all the variables included in the path model, the desired number of children shows the least effect. The other variables show greater importance in the following order: women’s ethnicity, women’s education, place of residence, and current contraceptive use.

We ask ourselves if the government is able to change fertility behavior in the country. The first question is: Can the government help in preventing further fertility decline and rehabilitating fertility among the part of the population where the fertility level is, in the long run, below the level of replacement? The second question is: Can the government help to accelerate fertility decline in order to slow down population growth in the part of the population where the fertility level is high – above the level of replacement? The answer is positive, but if family policies are considered as part of the labour market policies as well as gender policies, public child care for children of all ages and access to child care should be guaranteed as a social right of children. Namely, such a concept makes parents more relaxed in the sense that they care about their child or children, and the experience shows that there is a positive effect on increasing fertility, as is the case in northern countries in Europe. Besides, parental leave allows parents to care for their children without impairing their living standards or their employment. And on the whole, the support of families is based on providing social services rather than cash benefits (Esping-Andersen, 2002, 53).

In the present situation, Macedonia is making forward steps in the sense of managing these three aspects (facilitating mothers’ employment, alleviating mothers of their care work, and changing gender relations in care and employment) in the concept of promoting the family with more children. (Dragovikj, 2008)

Issues about the position of women (equality) are moving to the top of the global agenda – more in policy than in practice though.

Parental-leave schemes are a social right for both mothers and fathers, and are considered an important step towards changes in the family and in the employment sphere, but as such it does not exist in Macedonia. There is no

Fertility in the Republic of Macedonia 3

0

support for mothers in their care obligations through a scheme of various benefits (like the emphasis on job rotation and flexible labor market).

Strategies for demographic development in the Republic of Macedonia (MLSP, 2008) and activities in the field of family policies may have a modest effect in raising fertility levels, but in order to provide economically meaningful financial incentives, the country would require very substantial public expenditure. However, it may be expected that under prevailing social circumstances, anti-natal and pro-natal targets will not be equally realistic. Namely, setting and implementing a system of individual incentives with long-term pro-natal aims, under transitional characteristics of the country such as acute economic, political, and moral crises, chronic housing shortages, unemployment, and apathy may not accelerate the solving of the population balance in the country. So accelerating the fertility increase and at the same time preventing further fertility decline will undoubtedly be a difficult, but feasible, task that could be pursued through the promotion of family planning services and breaking down culturally induced segregation of women.

In general, it is expected that Macedonia will not diverge from the general fertility trend in the country. Based on the values of TFR, and in comparison to other countries, Macedonia is in the so-called “safety zone”. If we pay attention to the demographic future of Macedonia following the other specific aims in the strategy for demographic development for the period from 2008 to 2015, we can see that it refers to the increasing social capital, through higher quality education, improving access to and quality of health care for all age groups, improving social status of the population, and the program for increasing employment and reducing poverty, as well as the opportunity to influence the quality of life of the population, which is great, especially if it is realized in better economic situations.