Population Site: Articles and Papers
The Crisis of Population Knowledge in Turkey

Author: Frederic C. Shorter, Professor, Ataturk Institute (ATA), Bogazici University, 80815 Istanbul, Turkey. Correspondence E-mail: c/o support@hpntech.com

Publication: Frederic C. Shorter, "The Crisis of Population Knowledge in Turkey," New Perspectives on Turkey vol 12 (Spring) 1995: pp 1-31. Istanbul: Tarih Vakfi (Economic and Social History Foundation of Turkey). Reproduced with permission. Copyright (c) 1995 Frederic C. Shorter.

Contents:

Introduction

The main purpose of this paper is to present some basic information about the population of Turkey and its interpretation in a manner which readers of New Perspectives will find interesting for their purposes. One way to do this is to notice how certain population questions are handled in the public and scholarly discourse. Then, as a demographer and social scientist I may comment and hope to clarify wherever in my judgement that would be useful. Second is to present some of the main features of the macro-demography of Turkey as a short historical account. The account is unavoidably quantitative, but I hope digestible even by those who have aversions to numerical analysis. (footnote: For more statistical documentation a comprehensive assessment of Turkey's demographic structure and its development since 1923 is State Institute of Statistics (1995). The author participated in writing that study and draws on it as source material for the present article. The following persons deserve special recognition for bringing this valuable publication to completion: Professor Omer Gebizlioglu, Meryem Demirci, Fisun Sener, Bahar Uysal, and Orhan Baytok.)

When the truly large changes in the size and age structure of the population are seen, important questions about the effect of these changes on other fields arise, and it is interesting to think about them. Third, the urbanization of Turkey is one of the central and most important changes in the demographic structure of the society during the last 40 years, so its characteristics should be mentioned. There is a brief look at the effects on the demography of Istanbul.

Finally, it should become clear that the history of health (which affects death rates), of marriage and family reproduction (which affects birth rates), and of industrialization (which affects internal migration) are all interactive in systematic and demographically logical ways. Charts showing the history of these components of the demographic process are given separately. Their interactive outcome is also presented in pyramids and, later, as a historical chart of the national age structures of the population.

This paper looks primarily at the time sequences of the outcomes for key demographic variables. One of these -- birth rates -- receives most of the attention in demographic circles. There is nothing logically prior or necessary about this choice, since all aspects of the demographic system offer important opportunities for interpretation. Nevertheless, the focus of demographic inquiry is influenced by a common belief among the more policy-oriented demographers that high fertility is the root of most population problems, whether true or not in fact, and so it tends to receive the greatest attention. (footnote: The history of thought among American and European demographers has been very influential in two ways: by influencing policy formulations for the third-world at the centers of financial and intellectual power in the West, and by co-opting leading academics and policy planners from countries such as Turkey who study in the West and maintain intellectual contact thereafter. See for example, Dennis Hodgson (1983), and Finkle and McIntosh (1994).)

Selecting this topic -- birth rates -- a brief account is given in the last part of this paper of the theoretical frameworks that are employed to explain the decline of birth rates from very high to low levels in Turkey.

(return to contents)

Population momentum

In public and scholarly discourse, Turkey's population growth is often discussed, because it obviously has many implications for the country. The consequences of having a population of around 60 million, as at present, or a larger -- but how much larger? -- population in the future are discussed. One statement about population size made by a political leader is quoted, not to open a debate about points of view of different parties, but just to show the kind of reasoning that is implicit in some of the discussions. Then we may try to explain by demographic reasoning what the relationships actually are He says,

Our population which is approaching 65 million is not enough...population is the power by which to establish right in the world. These imitators of the West [family planners] try to decrease the population. Make at least four children!

-- Necmettin Erbakan, Refah Party Chair (Yeni Yüz Yil, 27 February 1995)

Turks at the present time (1995) are in fact making an average of 2.5 children per woman, and the figure is falling steadily. (footnote: This figure is the total fertility rate (TFR). It is a current measure. It shows how many live born children women are having at present. It is calculated by summing the age-specific birth rates of women for the ages of potential childbearing from age 15 to 49.)

Women, together with their spouses, will make only two children on average (plus a small margin to cover infant and early-age mortality) from the year 2005 onwards. This is according to the official population projection of the State Planning Organization (SPO) for the next Five-Year Plan. Concerning the size of the population, the projection shows that the population would reach 98 million before stabilizing around the mid-21st century. (footnote: A population projection is a forecast of future population size and age-sex structure based on assumptions about the future trend of fertility, mortality, and international migration. A mathematical model is used to make the projection. Such projections give reasonably good estimates for large countries such as Turkey for short time periods into the future; e.g., up to fifteen or twenty years. The assumptions can be violated by unexpected developments and this will cause errors, sometimes serious ones, for dates in the more distant future.

The State Institute of Statistics (1995: 66) made four alternative projections. They assume that the total fertility rate (TFR) will reach the so-called replacement level of 2.1 children by 2000 or 2005. The SPO selected the most conservative (highest) alternative of the four projections. Personally, my reading of the trends is that Turkey will reach a total fertility of 2.1 by the year 2000 at the latest. This alternative gives a maximum population of about 95 million.)

The reason why population continues to increase even after fertility falls to a level at which parents only replace themselves is due to the famous "population momentum," which is discussed immediately below. The SPO's "plan" for two- child families is simply a recognition of what people are doing. It is not an aggressive policy to persuade them to do anything different. Erbakan's admonition, on the other hand, asks them to change their actions and have four children. Others have endorsed the same reasoning, thinking that births at the rate of two per woman mean the population will grow no more.

Population momentum in Turkey can be visualized by looking at the following population pyramids for 1995, 2005, and 2050 which are based on the assumption that the average number of children per woman (and her husband) stabilizes at 2.1 children in 2005 (SPO assumption). I want to show why the population continues to increase in size when there is no excess fertility; i.e., that it is not necessary to have four children per family for the population to continue to grow for at least another 50 or so years.


    Population pyramids

           1995
  Males           Females

        *  75+  **
        * 70-74 **
      *** 65-69 ***
     **** 60-64 ****
     **** 55-59 ****
     **** 50-54 ****
    ***** 45-49 *****
  ******* 40-44 *******
   ******** 35-39 ********
 ********** 30-34 **********
 ********** 25-29 **********
   ************ 20-24 ************
  ************* 15-19 *************
 ************** 10-14 *************
  *************  5-9  *************
  *************  0-4  *************

   Size = 61.6 million


         2005
  Males           Females

      **  75+  ***
      ** 70-74 ***
     *** 65-69 ***
    **** 60-64 ****
   ***** 55-59 *****
   ******* 50-54 ******
  ******** 45-49 ********
********** 40-44 *********
********** 35-39 **********
  ************ 30-34 ***********
 ************* 25-29 *************
 ************* 20-24 *************
 ************* 15-19 *************
 ************* 10-14 *************
 *************  5-9  *************
 *************  0-4  *************

  Size = 71.7 million


          2050
   Males           Females

 *********  75+  *************
  ******** 70-74 *********
********** 65-69 ***********
   *********** 60-64 ***********
   *********** 55-59 ************
  ************ 50-54 ************
  ************ 45-49 ************
  ************ 40-44 ************
 ************* 35-39 ************
 ************* 30-34 *************
 ************* 25-29 *************
 ************* 20-24 *************
 ************* 15-19 *************
 ************* 10-14 *************
 *************  5-9  *************
 *************  0-4  *************

  Size = 98.2 million

Note: Each symbol (*) represents
  approximately 255,000 individuals.
      

When there is no excess fertility above approximately two children per woman, the number of individuals added at the bottom of the pyramid (0-4) stabilizes. Notice that the bottom of the pyramid is the same size in all three pyramids, including the last one in 2050. Every five years, the individuals on one age level of the pyramid move up one level until they disappear from the pyramid due to death (or out-migration).

As individuals get older they face the risk of death according to the age-specific mortality conditions of the time. The older people in the 1995 pyramid were born long ago when the expectation of life at birth was a good deal less than it is now, so fewer survived. (footnote: The expectation of life at birth is the average number of years that a newborn babies may expect to live before dying if health conditions (age-specific death rates) do not change during their lifetimes. Of course, reductions in age-specific death rates historically over time give to those who are still surviving some additional years of life expectancy. Probabilities of dying at older ages are to a great extent, but not entirely, independent of the health conditions that persons faced in their younger years.) Furthermore, the population was smaller and hence births were fewer. So fewer people were added and survived to reach the upper part of the pyramid; i.e., it is not very fat.

However, as time passes after 1995, people will get older and more of them will survive with the high expectation of life -- currently about 65 years. Life expectancy will continue to increase, so the projections assume it will reach the low 70s. The basic reason why the total population keeps on growing is that people get older without dying until rather high ages are reached. The fact of "not dying" adds to the width of the pyramid at older ages as time passes. Compare the last pyramid for 2050 with the one for 2005. All of the increase -- population momentum -- is in the age groups above 30 years of age.

The last pyramid shows the age and sex structure, and the population size, after the population momentum is basically exhausted. This pyramid will go on being reproduced year after year after the population stabilizes. The projection foresees a constant population size and structure from the mid-21st century onwards, unless fertility rates fall or rise, or mortality rates get better or worse, or international migration makes significant additions or subtractions of individuals.

If Erbakan -- or for that matter any change in the agency of political authorities or in local culture -- were able to persuade every family to have four children, and all women to marry, the pyramid would get wider and wider at the bottom. Population growth would be much faster without any tendency to stop growing. By the year 2050, a population projection based on the four-child assumption shows that Turkey would have 122 million people with a growth rate of 2.4 per cent per year, which would continue indefinitely thereafter into the future.

At present (1995), the rate of growth of the national population is 1.6 per cent per annum. (footnote: Excluding international migration about which information is scarce and incomplete.) It is projected to be 1.4 per cent by the year 2000, and to decline thereafter to approximately 0.0 (zero) at the middle of the next century. These figures about the growth rate are often confused in the public discourse with the total fertility rate -- currently 2.5 children per woman. A growth rate of 1.6 per cent per year for the total population, and a fertility rate of 2.5 children at the present time are completely different things. The latter helps to determine population growth rates, but is certainly not the same as we have seen.

(return to contents)

Is high fertility a problem?

At the highest levels of the state, and particularly in the State Planning Organization and the Ministry of Health, the doctrine since the mid-1960s was that fertility is too high in Turkey and should be reduced by family planning. The first law to this effect was passed in 1965, and another was enacted in 1983. No very strong steps were ever taken to translate the policy stated in the law into state actions (see discussion later in the paper), but the view that fertility is too high and is a serious problem for Turkey was upheld in the discourses of the political, intellectual, and public cadres. There was dissent and often attempts simply not to talk about the problem, Today the basic doctrine persists, though there are other views as we have seen above.

At the intersection of international and national policy-making institutions, advocacy of family planning as a way to face the perceived problem of high population growth rates in Turkey has a long history. An example from recent times is given next.

At a ceremony in Geneva on June 14, 1994, the Secretary-General of the United Nations, Dr. Boutros Boutros-Ghali, presented Mr. Vehbi Koc in his capacity as President of the Turkish Family Health and Planning Foundation, the annual U.N. Population Award for "the most outstanding contribution to the awareness of population questions or to their solutions." The foundation, established in 1985, was cited as "a model of what can be achieved by the private sector when it cooperates by leading and augmenting government programs." (footnote: Documentation and press release at the award ceremony.)

Mr. Koc, a leading industrialist, has advocated the ideal of the "two-child" family ever since the 1970s. The TAP VAKFI (Turkish acronym) which received the prize was a newcomer, and an innovative one. It launched a mass media (TV) campaign on behalf of family planning in 1988, the first ever in Turkey. The foundation also produced a prize-winning film, Berdel, which depicted the folly of men who insist on fathering many children. In addition, the foundation organized a consortium of pharmaceutical companies, placed spots on TV, and promoted the sale of oral contraceptives and condoms in pharmacies all over Turkey. Two model family planning clinics were also opened.

At that time, 1988, total fertility per woman had already fallen from 6.6 children in the 1950s to 3.3 children. By 1994, when the award was made, total fertility was 2.7, and in Istanbul where the foundation had its headquarters, fertility was about 2.0. We think that these measurements are reasonably accurate. (footnote: All measurements in the present paper, except where noted, are from State Institute of Statistics (1995). Readers should look there for documentation of sources and methods.)

The family planners' goals were to persuade more people to use contraceptives and thus to lower fertility. They hoped to see more use of methods of contraception which they describe as "modern methods" in contrast to the principal method used in Turkey, which is withdrawal (coitus interruptus) and is described as "traditional." The results obtained by TAP VAKFI and the participating pharmaceutical companies as a result of their campaigns in the late 1980s and early 1990s are instructive.

There are national surveys every five years in Turkey to measure the prevalence and types of contraceptive practice along with some other subjects related mostly to health. According to these surveys there was no increase (actually a decline) in the use of pills and condoms between 1988 and 1993 (Hacettepe, 1994: 38). The pharmaceutical companies involved in the campaign had higher sales, however. Perhaps they succeeded in gaining a higher market share, or the Hacettepe surveys were not sufficiently accurate.

The most interesting point is that the total fertility rate (TFR) continued to decline. Thus, the contest between the leader of the Refah Party and the family planners is for nothing. The former will get a larger population without the need to do anything about convincing people to have more or less children, and the latter need not address the quantitative problem though they might contribute new alternative means of control and information about their birth control problems that men and women could use. The quantitative goal of the family planners is more or less already reached, at least as an average for Turkey, admitting that some groups have higher and other groups lower fertility.

There is, however, another story. A new public discourse that transforms the family-planning orientation to concerns with the reproductive health of women, mostly, and men, to a lesser extent, begins to be seen. The appropriate question is, how could women and men increase their knowledge and experience of the potentialities and options so far as their sexuality is concerned, and attain a better quality of management for their reproductive health. Could the health services help them in this regard? Finally, it should be asked how marginal groups, such as lower-class social groups and residents of remote regions, could be offered support for the control of fertility, which they are in most instances already doing with some privately-achieved results.

These subjects have not so far attracted much attention from the various actors, probably because of their previous preoccupation with the quantity (prevalence) of contraception, which is supported by the internal culture of the state and private institutions. (footnote: Projects organized in recent years in cooperation with international organizations such as the U.N. Fund for Population, the U.S. Agency for International Development, and the European Union, are based on documents that say more family planning (read "medical contraception") is urgently needed because the population is growing at a high rate; 2.5 per cent per annum is a figure often used. This agrees with the diagnosis of the international family planning "industry," and helped to bring projects and funds to Turkey. The point of view does not go away easily, particularly since it is a formula applied to many other third-world countries as well.)

An exercise of agency by individuals who have some power inside the system could redirect attention toward quality, which is a one-word concept to embrace all the new ideas. Some small beginnings are seen in the Ministry of Health, State Planning Organization, and the clinics or foundations staffed by professional family planners. Scholarly attention to these issues is also seen, but it too is in its infancy. (footnote: An exception is the very informative study of the quality abortion services in Istanbul, which also provides incidental findings about other aspects of birth control (Bulut and Toubia, 1994).)

The U. N. Cairo Conference on Population and Development in 1994 offered support for the new emphasis on reproductive health as a broader concept, though family planning remains as a sub-category within reproductive health. Here again is an illustration of the intersection of international and national institutions at the political level. Much will be said and written, but to what effect on Turkish lay culture, and the day to day lives of people remains to be seen.

(return to contents)

A short history of macro demographic changes in fertility, mortality, and age structure

A population grows by virtue of its own history of age structure in the past and the additions or subtractions by births, deaths, and international migrations (to or from Turkey). As we have seen, the history of the population, as well as the actions of people with respect to reproduction and migration, plays a very important role in determining what will happen to the structure in the future. To understand today's situation, therefore, demography must look at its own history.

In 1923 the age and sex structure of the population bore the scars of a long period of war and privation. It was short of young adults, particularly men. The normal processes of population renewal had been thwarted by both depressed fertility and high infant and early childhood mortality during many years before 1923. The first thing to happen demographically was a rapid rise in fertility. Peace was a wonderful thing. Existing families could be reunited, new families formed, and demographic recovery could proceed at the family and national level. (footnote: For information about this period and an interpretation, see Shorter (1985).)

It was 30 years later, in the 1950s, that the famous fertility decline began, gaining speed in the 1970s, and continuing up to the present time. Therefore, the main demographic change that created the high rate of population growth in Turkey, reaching almost three per cent per year in the 1950s, was improved health and a rising expectation of life; i.e., reduced death rates, not increased birth rates, though they made a small contribution.

There was an incredibly large increase in the average expectation of life at birth. It started from a measured level of 35 years before World War II.(footnote: The figure refers to 1935-40. Good measurement of mortality before 1935 does not exist. It is simply assumed that death rates were even higher in earlier years.) The figure then rose by 30 years up to 1989. In that year a national survey found that the expectation of life had reached 63 years for males and 66 years for females (SIS, 1991:101). Death rates for infants and two or three-year olds play a major role in the expectation of life at birth. In those early years, more than one-third of children were eliminated by death in the first years of life; today the figure is not more than five per cent. When so many children die, their years of life contribute very little to the calculation of the average expectation of remaining years of life at birth. So the reduction of infant and early childhood mortality played a major role in raising the average expectation of life at birth. There were other factors as well. For example, the decrease in childbearing during the last 30 years meant that many fewer women died in childbirth or complications of pregnancy as compared with the early years. Improvements in diet, living conditions, and medical assistance also contributed.

A small chart is given to show the changes in infant mortality (IMR) which signify the general change in infant and early childhood mortality as a whole. A chart for the rise and then decline of the total fertility rate (TFR) is also given. (footnote: The sources of data are the same as those given earlier (SIS, 1985). For infant mortality prior to 1965, see Shorter and Macura (1982:64))

We will discuss the fertility decline in more detail later. The present situation is one of low fertility and low mortality. The broken lines indicate what might be expected in the immediate future. These two components, mortality and fertility, are the main determinants of the population's growth and the age structure over time. International migration plays a role, but it is neither well documented nor is it very important relative to the other two factors. So the overall development of the national population is based to a great extent on these two factors.

Total fertility per woman

 
Infant mortality per thousand births

The fertility and mortality trends reflect momentous changes in the society, and they have important consequences as we shall discuss below. The rise of fertility from 1923 up to its peak in the mid-1930s, which was then sustained with only a small decline during World War II, added children to the population at the 0-4 level of the pyramid. The cohorts of new population (births) had to grow up over many years, following one after the other and rising higher and higher in the pyramid, before the pyramid would lose the scars of the initial Turkish population in 1923. This was substantially accomplished by around the 1960s.

Meanwhile, the decline of mortality (improvement in survival rates at all ages) caused the population to grow rapidly. Population size grew from 13 million in 1923 to 31 million in 1965, and 61 million in 1995. (footnote: Since the last census was held in 1990, the figure for 1995 is an estimate made by population projection.) During the 1950s and 1960s the rate of growth of the national population reached its highest level of 2.8 per cent per year.

(return to contents)

Marriage

Associated with the changes in fertility there were important changes in the process of marriage and in the ages at which people (first) marry. It seems necessary to present some information on this point, even though it is well known informally in Turkey. Scholars in particular have been misled at times and published what we would call "counter intuitive" information on this subject.

An American anthropologist, searching for information on this point to relate to her story of how marriage takes place in Istanbul and how women are brought into the labor force, went to normal demographic sources and found a figure for Turkey. (footnote: Admittedly, I select a negative example even though some more informed writings also exist. I do this to show that no "safety net" for authors seems to exist; do reviewers or editors ask questions?) She writes,

...more recent statistics (1988) show that the mean age of marriage for women (in Turkey) has risen to eighteen. (White, 1994:41)

The author of this statement obtained the figure from a report by the Hacettepe Institute of Population Studies (1989: 29-30) referring to their national 1988 survey. Unfortunately, there was an error in the methodology for the calculation. The same methodological error was repeated five years later in the 1993 survey report with the same consequences (Hacettepe, 1994: 62). This was very unfortunate, because other data in the Hacettepe report are properly prepared and presented. Even if one does not have an inside knowledge of the demographic literature, general knowledge of social change in Turkey should make one doubt and ask questions.(footnote: Looking closely at the 1993 report, one discovers a table on the marital status of women (Table 6.1) which could be used to calculate a mean age of marriage of 22.4 years. However, the calculation was not made by the Hacettepe demographers. This figure is plotted with a + in the chart.)

The "true" story, calculated from the censuses, is shown in the chart. A preliminary (partial) interpretation of the trends shown there would go as follows: The decline in the age index over the years prior to the 1950s was associated with the recovery of family life after men were released from military service in the 1920s and while labor-intensive agriculture remained the main motor of economic advance for families and the country. That was a period of rising and then sustained high fertility.

The blip for males during the 1940s was due to wartime mobilization of men which took them temporarily out of the marriage market. Thereafter, the rise of the index by four years between the 1950s and the present time signify several changes in the marriage process: the decline in number of early arranged marriages, more time spent on education and negotiation by young persons of their own marriages, and recognition by young persons that economic advance depends on taking time to position one's self to work in urban industrial environments.

(return to contents)

Structural changes in the national population

The result of all these changes that we have mentioned separately for statistical convenience is a national population whose structure changed greatly over time. A chart to show the history of how three age groups of the population increased absolutely and in relation to each other over these years, beginning from the 1935 census, is given next. The chart also includes a population projection from the present time to the mid-21st century. This puts in perspective the changes in population structure (age and sex) over this long period.

Youth, producer, and elderly population,
1935-1990, with a projection to 2050
(millions of population)


solid line: Producers (20-54)
dashed line: Youth (0-14)
dotted line: Elderly (65+)

The producer group is defined as men and women aged 20-54, including both sexes because women generally work in domestic tasks if not market activities whether recognized in censuses and national accounts or not. The youth are those under age 15 and the elderly are those aged 65 and over. These age categories are a way of drawing attention to the changing structure. They summarize what is also shown in pyramids, but on a historical continuum. They also focus attention on certain periods during the life course of individuals as young people, producers, and elderly.

Prior to the 1980s, an elderly population scarcely existed, because it took many years for the new additions to the population that began in the 1920s to become old enough to enter this age range. With higher life expectancy and the passage of time, this part of the population will become increasingly prominent and a focus for many of the country's population questions. The role of older people in politics, in the labor force to the extent institutional conditions permit them to work, and their demands on private and public health care systems will be important. The question of how to support themselves, through state security institutions or by private family arrangements, will have to be addressed.

Until the 1980s, there was one adult producer (typically a father or mother) for each young person or child under the age of 15. Two bits of demographic history are now causing this relationship to change dramatically; soon there will be two adult producers per person under age 15. Declining fertility has reached the point where the lower age groups are more or less stationary in size; they are being reproduced but not increased in number. The producer group, on the other hand, is growing rapidly. Notice how the curve shows acceleration in the rate, followed by a slowing down in later years.

The potential of parents to make economic and social investments in their offspring has increased. The demographic figures alone show a doubling of this capacity. Any increase in family real incomes adds even more to the potential to invest resources in each child. This bodes well for the future quality of the Turkish population. Parents today, far more than "yesterday," are demanding educational opportunity for their children and are paying the costs of giving them many other options in life in addition. In fact, everyone knows this at the micro demographic level by personal experience and the local culture for children they see around them. Families have re-oriented themselves to lower levels of reproduction with qualitative objectives for themselves and the next generation in mind. We point out that this is a consequence, partly not wholly, of the new relationships between parent and child generations made possible by the demography.

The labor force, on the other hand, is passing through a particularly difficult period if we look at it from the perspective of individuals who require employment. The reason is that this age group is growing so fast. The yearly rate of growth has been rising for some time and reached 3.0 per cent per annum in 1993. This is a peak rate. It will start declining soon, but only slowly, reaching 2.0 per cent by the year 2005. Notice that this age group is experiencing the time-lagged effects of rapid growth at young ages earlier in history when fertility was still high and death rates were declining. It is important not to look at the general rate of growth in the total population which is only about half as fast. The consequences for individuals is that they face more competition in labor markets and can be exploited by tough wage bargaining policies. During the Ozal period (1980s) this is precisely what was happening and part of the reason was demographic.

From the perspective of capitalist enterprises the abundance of labor is a benefit. It was no doubt one of the reasons for the remarkable rates of increase in economic output in the 1980s. For any system of production that is relatively labor intensive, this time period is an opportunity from the perspective of employers; e.g., blue jeans, textile manufactures, tourism, transport, and so on. The 1990s are more problematic, because macro economic policies can support or interfere with the institutional arrangements necessary to make use of the labor option. Turkey's economic troubles of the 1990s seem to have prevented full exploitation of this abundance of labor.

From the state's perspective, labor abundance is also a benefit. It can support an export boom as in the 1980s, or army recruitment as in the 1990s, or both in differing degrees. The three perspectives of labor, employers, and state, will eventually face a change, since the high growth rates of the labor force finally decline, even becoming 0.0 (zero) around 2025. As that change occurs, slowly as is typical with demographic change, a move toward quality and less labor intensive occupations and systems of production should occur. The stimulus just discussed was a one-time package; it won't come again. (footnote: Barlow (1994) notices, as do we, that this type of demographic stimulus weakens and disappears.)

Turkey is not alone in experiencing these opportunities and problems -- the perspective tells which is opportunity and which is problem. The economist, Dowrick (1992), attributes high growth rates in East and Southeast Asia to "declining birth rates (that) provide a temporary but substantial stimulus through this labor force effect." It happened earlier in that region, because the demographic conditions occurred earlier.

A dark side of the picture is the slow progress in raising educational levels in Turkey. This is a drag on the improvement of labor force skills and experience. The slow progress of educational achievement is also related to demographic factors. Until now, demographic growth presented the educational system with rising numbers of persons to educate, and there was a preoccupation with expansion of scale rather than of quality.

The following statistics are useful to assess the progress of education among individuals entering the labor force. They refer to persons when they are aged 20- 24, by which time they should potentially reach at least a high school education level. In 1990, only 28 per cent of males had reached such a level, and 19 per cent of females. Of course, this was better than the situation 20 years earlier in 1970 when the figures were 12 and 6 per cent respectively. Thus, a labor force most of which was poorly educated -- primary school was the typical level -- during those earlier years will improve as it is renewed by young cohorts that enter the labor force with higher educational qualifications.

(return to contents)

Urbanization and fertility in the cities

Another aspect of demographic structure is the spatial distribution of population. This is discussed within the framework of internal migration and urbanization.

The early projects of the state to industrialize Turkey focussed on certain cities and industries. In Istanbul, for example, workers were needed to staff large-scale factories in glass, leather, alcohol, and tobacco. This was migration without any serious "push" forces from the countryside. The size of the urban sector increased gradually until the 1950s. The elasticity of expansion of agriculture eventually could not absorb the rate of population growth, and by comparison with the cities, strong economic inducements to migrate came into existence. The cities "exploded" from the 1950s onward. A table from the SIS report (1995) is included below to show this. The last column shows the well-known fact that Turkey is becoming a dominantly urban society. The proportion living in urban places is almost certain to rise to 65 or 70 per cent by the year 2000.

The rural population is no longer increasing. From now on, it seems that all natural increase (births minus deaths) in the rural areas will be compensated by transfers of rural people to urban places. This means that although births may exceed deaths in the villages, individuals in the age range that migrates -- late teens and young adults -- make enough moves to the cities to offset the natural increase. This also has a limiting effect on the growth of the villages by births, because the migrants usually have their children after marriage and arrival in the cities. Primary schools are being abandoned or becoming half empty in the villages and small towns. Turkey even has abandoned "ghost" towns and villages. (footnote: We do not refer to areas of the country where security factors are responsible for internal migration. About this subject, there is little acceptable demographic information. Moves prior to 1990 are reflected, however, in the 1985-1990 census comparisons. See SIS (1985: Table 5-4.)) Meanwhile, the educational and health facilities of the cities that receive migrants, especially those with high in-migration rates, are unable to cope with the increased demands on their services. (footnote: For example, the ratio of net migration to the existing population in large cities (over 500,000) is highest in the industrial city of Bursa with approximately 30 migrants per thousand population per year, followed by Istanbul with 23 migrants per thousand population per year. The government city of Ankara has a rather low net migration rate of only 5 migrants per thousand population per year.)

Urban and rural population, and percent urban, 1927-1990
 
Census Total
Turkey
Thousand
Urban
10,000+
Thousand
Rural
<10,000
Thousand
Percent
Urban
1927 13,648 2,236 11,412 16.4
1935 16,158 2,735 13,423 16.9
1940 17,821 3,203 14,618 18.0
1945 18,790 3,442 15,348 18.3
1950 20,947 3,782 17,165 18.1
1955 24,065 5,425 18,640 22.5
1960 27,755 7,308 20,447 26.3
1965 31,391 9,383 22,008 29.9
1970 35,605 12,754 22,851 35.8
1975 40,348 16,707 23,641 41.4
1980 44,737 20,330 24,407 45.4
1985 50,664 25,890 24,774 51.1
1990 56,473 31,805 24,668 56.3

Population is transferred from rural to urban by (1) actual physical movement of population, and (2) by reclassification when a small place just keeps on growing in size and crosses the boundary definition of 10,000. Pure migration (1) was approximately 670,000 per year during the inter-census period of 1985-1990. At the beginning of the massive internal redistribution from rural to urban which commenced in 1950s, it was 240,000 per year. The figure has been increasing, but seems to be near a maximum at present, after which it will most likely decline. The migration figures are net -- from rural to urban minus from urban to rural.


solid line: Urban
dotted line: Turkey
dashed line: Rural

The largest cities of Turkey, and perhaps smaller ones as well -- we do not know -- already had controlled fertility and low levels of reproduction before the 1950s. One effect of the redistribution of population from agricultural villages and towns to the industrializing cities was to move generations of young adults into the controlled-fertility environment. These generations adopted the lower fertility. We may illustrate by following fertility in Istanbul over time.

(return to contents)

Istanbul

Soon after the turn of the century, the Istanbul total fertility rate, was 3.85 children (Duben and Behar, 1991: 162). (footnote: These demographic rates were computed by Duben and Behar for the Muslims only, because they were working with the Muslim family population records (nüfus defteri). Of course, our figures after 1923 include everyone whether Muslim or not.)

If we deduct deaths between birth and maturity of the generation, net reproduction may have been higher than replacement, but not by a very wide margin (mortality information is lacking). In the 1940s, we find a fertility rate of 2.41 children (Shorter and Macura, 1982: 51). Again deducting child mortality, Istanbul must have been a city that barely replaced its own population by fertility. The sequence of Istanbul fertility figures from then up to the 1990s shows a slight rise toward 3.0, but never reaching that figure, a plateau, and then a decline starting in the 1980s that takes Istanbul down to a replacement level or below by the mid-1990s. No doubt, the rise and then fall in Istanbul was due to the absorption of large numbers of first-generation migrants after 1950. The more recent waves of migrants are just as large, but they are generations of children from parent families who were already controlling fertility.

It is a simplification, but one can say that Istanbul in this century had a population that did not replace itself by its own fertility (minus mortality). Its own population made a negligible or no contribution to its own growth. The city grew by virtue of migrants who arrived with children not yet born and added whole families (the new generation) to the city. Thus, when we note that the net (in minus out) flow of migrants to Istanbul is an average of 133,000 per year, (footnote: The figure refers to 1985-1990. During the previous five years it was 120,000 per year.) we should think of their adding at least twice that number to the city because their fertility will be high enough to replace themselves in the next generation. Without this process, Istanbul could not have grown from less than one million in 1950 to nearly eight million by 1995.

A graphic way to see how migrants and children are introduced into the population is to examine the population pyramid for Istanbul in 1995. (footnote: Based on 1990 with a short five-year projection to 1995. As shown in SIS (1995:Chapter 5).)

The fat waist for the young adult ages is due to net in-migration. At the bottom, the first three five-year age groups are the children produced by the fertility of the population, which is clearly not sufficient to replace the adult parents above them, even though many of these adults have not yet finished bearing their children. The upper part of the pyramid is slender, because it receives little migration and will be enlarged only as the young and middle-aged adults grow older, without dying, as time passes.

There is some momentum of population growth in Istanbul in the same way that there is in the national population (discussed earlier). However, if all migration were to stop today, momentum alone would not carry the population size of Istanbul to more than about ten million. After that, due to below-replacement fertility, Istanbul would decline!


 Istanbul's population pyramid, 1995           
   
    Males       Females
   
        *  75+  **
        * 70-74 **
       ** 65-69 ***
      *** 60-64 ***
      *** 55-59 ****
     **** 50-54 ****
   ****** 45-49 ******
   ******** 40-44 *******
 ********** 35-39 *********
  ************* 30-34 ***********
  ************* 25-29 *************
*************** 20-24 ************
 ************** 15-19 ***********
   ************ 10-14 ***********
***********  5-9  ***********
***********  0-4  ***********
   
   Size =  7,744,000.
   
Each symbol represents approximately
32,000 individuals.    
      

Since the cities had controlled fertility -- not Istanbul alone -- the growing proportion urban, which we call urbanization, brought more and more of the population into economic and social contexts where families needed to control fertility in order to survive. This has contributed in the Turkish case to the national fertility decline. We would emphasize that Turkey had its own particular character, with urban centers of low fertility since at least the beginning of the century.

(return to contents)

How did these changes in demographic structure happen?

When demographers answer the question, how did all these changes in demographic structure happen, it seems rather obvious. Given the changes over time in fertility and mortality shown in the charts, and leaving aside international migration, the resulting increase of population and the chart of age groups shown above are logical numerical results. Even the amount of urbanization is not very problematic. Given such large increases in urban-based systems of economic production, it simply had to happen. If, however, we ask for explanations of the improvement in health standards and expectation of life, that is more problematic. Then, if we ask how fertility happened to rise and then fall to such low levels, that too will be more problematic. Historical analysis with appropriate theoretical perspectives is needed. We have already gone some distance in explaining, at least demographically, how the fertility decline took place. Now we want to go into the matter more thoroughly.

There are three theoretical frameworks that are applied to the fertility question. The answers from one do not preclude the other, but readers of different theoretical persuasions concerning social change will find one or another more convincing and complete. The three are: first, the bio-demographic perspective; second, the demographic transition or modernization perspective; and third, a collection of cultural perspectives.

(return to contents)

The bio-demographic perspective

From the bio-demographic perspective, the decline of fertility is interpreted as a consequence of the application of procedures of contraception (protection against pregnancy) and abortion (intentional termination of pregnancy). If a policy objective of reducing fertility is adopted and questions are asked about how to accomplish this result, the bio-demographic answer is to inform people about how to do it, motivate them to use methods, and provide whatever medical support may be necessary. Social questions are not at the center of this perspective.

The following table summarizes the information about contraception that is needed by the bio-demographic perspective to account for the fertility decline in Turkey. There are some other factors, such as breastfeeding practices, the duration of exposure to the risk of pregnancy by co-habitation, and termination of pregnancies by abortion. We will only discuss the last, abortion, below.

Methods of protection used by couples in Turkey at an early year of the fertility decline (1963) and thirty years later (1993) Thousands of couples (Per cent of all users)

Methods Used 1963 1993
Male dependent:
    Withdrawal 495 (38) 2,730 (42)
    Condom 205 (16) 688 (11)
    Sub-total male dependent 700 (53) 3,418 (52)
Female dependent:
    IUD 0 ( 0) 1,959 (30)
    Hormonal (mainly pill) 48 ( 4) 521 ( 8)
    Sterilization 0 ( 0) 302 ( 5)
    Diaphragm/foam/jelly 67 ( 5) 125 ( 2)
    Douche, potion, blockages 505 (38) 198 ( 3)
    Sub-total female dependent 619 (47) 3,105 (48)
All couples using a method 1,319 (100) 6,524 (100)
Not protecting at time of survey (e.g., pregnant, willing to become pregnant, or any other reason) 3,443 3,897
All couples (a) 4,762 10,421
Per cent using a method of protection 27.7 62.6

(a) Couples are counted by looking at the number of women who are married (under age 45 in the 1963 data and under age 50 in the 1993 data). We use interpolation between censuses and projection where necessary. The sample surveys show the percentages of couples using each method, which is multiplied by the total number of couples to obtain the figures above.

Sources: 1963 Survey (Ozbay and Shorter, 1970: 5); 1993 Survey (Hacettepe Institute of Population Studies, 1994: 38).

Notice that medical methods of control were simply unavailable and unused at the beginning of the fertility decline (1963 figures). By 1993 they are more important: IUDs and Pills together are 38 per cent of all methods used, but the so-called traditional, or male, methods continue to be the dominant mode; i.e., 52 per cent.

Abortion is also known to have been common for a long time in Turkey, but quantification was not satisfactory. At present, survey data show that 18 per cent of pregnancies are terminated by induced abortion (Hacettepe, 1994: 52-53). The percentage is declining slightly over time, which could be due to many factors, but one of them might be more success at avoiding pregnancy by contraception. Abortion was legalized in 1983, but the opinion of medical providers is that the change in the law had no major influence on its use. The main gain was to mandate the establishment of abortion services in social security and government hospitals and clinics. The law reinforced the power of men in decisions about birth control by requiring a signed consent from the husband for any married woman who requests an abortion.

The main question raised by the bio-demographic perspective is how to interpret the fact that so-called male methods of contraception, withdrawal and condoms, are the most important, both at the beginning and end of the fertility decline. (footnote: The importance of such methods in other countries and contexts is discussed in Santow (1993).)

At the beginning it is not so surprising; little else was possible and fertility had to be controlled. At the present time, however, it is more problematic. If technology is the way to bring about change, why is it not being used more?

All methods generally involve female cooperation no matter whether the application is related medically or mechanically to either the male or female body. The biology of a method does not really tell us "who is in charge," or who (husband or wife or both) is the one acting to keep the number of births to a socially accepted small number. That is a subject which should be addressed by examining the social construction of gender roles and birth control, and by placing the construction in an appropriate context. It belongs to the domain of cultural analysis to which we will come below.

(return to contents)

The demographic transition or modernization perspective

The demographic transition perspective is a close relative of modernization theory. It derives its principal ideas from that discourse. The starting point is a "traditional" society with the implication of static, unchanging fertility behavior. Already, that is a troublesome point for Turkey, because low fertility was established in some places such as Istanbul very early, and in others it came much later. Furthermore, there was the period of rising fertility during the first 30 years or so of the Republic's own modernization policies. Though the extent and speed of their penetration to all corners of Turkey may be discussed, the establishlment of many formal institutions, industrial ventures, Latinization of the language, new civil laws, transport, and so on certainly affected daily life. According to transition theory, modernization is supposed to bring fertility down, not raise it. This is a problem in the interpretation of fertility declines in other countries as well. (footnote: There is a substantial literature that documents the disappointment of research scholars with the modernization approach. See for example, Knodel and van de Walle (1979), Watkins (1986), Coale (1969), McNicoll (1980, 1994), Greenhalgh (1994).)

Nevertheless, there might be some possibilities for using this framework for the period following the 1950s. As modernization proceeds, social-structural incentives to have fewer children come into force. These are such contextual conditions for families as the following (we merely list them): changes from family labor to wage labor in systems of production; educational requirements for employment and compulsory education in any event, both of which put up the time and money cost of children; improvements in the status of women, which is supposed to put them in position to exercise a "natural" tendency to want fewer children; improved health for children so that fewer need to be borne, expansion of options for women and men that are alternatives to having children, such as consumer goods and better housing; and communication of ideas and means for practicing birth control.

Modernization is also supposed to create the technological means for birth control, which will be provided by the medical system that is undergoing modernization. This last point is particularly for third-world countries, because it did not really exist for the countries that underwent their transitions much earlier. Basically, all these changes raise the costs of having children, lessen the immediate monetary advantages, and surround people with options for a way of life that involves smaller families and non-familial commitments of time and money.

In some degree, all of these things have happened in Turkey, although the one about the status of women is problematic. So there is a litany of factors that could explain the fertility decline. The difficulty with this perspective is that it sees adults as individuals making choices and reacting to changes in the structural environment without any reference to history or their local culture. It doesn't really link up some factors closely and exclude others. Nor does it help very much to explain the timing of the fertility decline. All the connections are a bit loose. Yet, for some social scientists, one or more packages of these factors seems sufficient to handle the question of interpretation.

One of the tenants of transition theory is that demographic behavior should converge with a kind of homogenizing result. In one sense, this is found demographically. Fertility in the cities cannot go much lower than it is, so the declines elsewhere in the country do narrow the differences. By the end of the 1960s, fertility was declining everywhere in Turkey. It is sometimes thought that the East would be an important exception, so we should look at information on that region in particular. Does it fit into the homogenization expectation of transition theory or not?

The initial levels of total fertility in Eastern Turkey were very high. Figures for 1960 are eight to nine children in provinces of the southeast (Shorter and Macura, 1982: 53). Good measurements at a province level of disaggregation are not available since then, but satisfactory regional measurements are. For the Eastern region of Turkey, defined to include the southeast and other eastern Anatolian provinces including Sivas, the figures are shown in the box. The macro demographic data leave little doubt that the decline had begun in the East by the 1970s.

Total Fertility: East
1960 8.3
1978 6.9
1983 6.7
1988 5.6
1993 4.4

The other four regions of Turkey, defined by the State Institute of Statistics as south, central, north, and west, have each one experienced declines of 50 to 60 per cent, which is a bit more than the decline (proportionately) for the East. However, there does not seem to me to be enough difference in the speed of decline to construct a special theory that would say that the East is a special case. No doubt in all the regions the way in which the declines have taken place differ in details, but transition theory is not much concerned with this kind of variation. Social scientists who apply one or more of the cultural perspectives are more likely to be concerned with such variation.

(return to contents)

History, culture, and agency

Many scholars of fertility declines have suggested one or more different ways to explain the decline from the top down as a problem in the political economy of fertility. (footnote: Susan Greenhalgh (1990) reviews these approaches in her paper on anthropological contributions to the political economy approach.)

The Turkish state advocated high birth rates in the early decades of the Republic, and the people of the countryside -- less so in the cities -- had high birth rates. The state's laws made it a criminal offense to import, produce, or distribute contraceptives, or to distribute information about methods. Abortion was also a criminal offense. Though these were the declarations and the laws, people could control their fertility by their own means, sometimes dangerously by self-induced abortions. (footnote: Dr. Pakize Tarzi, the famous gynecologist, writes about her visits to villages in the early years of the Republic where she found self-induced abortions to be widespread (Tarzi, 1992: 38)) Money could also buy competent medical abortions. So long as families wanted to have five or even eight children, as many did in rural and small-town Turkey during those early Republican years - - to fill in the gaps in their families and to have adequate labor power for the systems of production -- the state and people appeared to be in agreement.

It would be difficult, I think, to argue that the state adopted a policy, took action through the means at its command, and somehow compelled the population to act in agreement. It would be more convincing, perhaps, to say that the state and the population arrived at similar conclusions about the appropriate level of childbearing, and both acted with the same end in view -- a consensual rather than top-down result.

After World War II, the consensus began to come apart in a large way. It already did not exist with respect to urban populations that were controlling fertility. However, now there were many changes in the society, socially, economically and in where people lived -- urbanization picked up speed. People did what anthropologists say; they reflexively monitored their local cultures and rationalized a continuous flow of conduct that brought them to more and more birth control. So-called "traditional" birth control, but not tradition in the sense of a static unchanging pattern of action, played a major role as we have seen. People were acting to reach goals that they chose, mobilizing and applying the resources within their command.

It is particularly interesting that the male methods which predominated at the beginning continue to be the most important now. They have been instruments in the process of achieving a low level of fertility. The principal modification of birth control over this 30-year period is that the so-called female methods have become more medicalized (mainly IUDs and Pills instead of vaginal potions, douches, and blockages). As the medical establishment, including pharmacies, became better equipped to offer the so-called modern methods, there was some uptake.

The turning point in state policy came in 1965 when a law was enacted that directed the Ministry of Health to facilitate voluntary family planning by investigating and approving methods, and offering methods through its own clinics and hospitals. The state opened some of the doors that previously prevented a flow of contraceptives, but in fact did very little through its own facilities to encourage the use of medical methods of control (Behar, 1980). In 1983 the state took a further step and removed the criminal sanctions against abortion. It also ordered more training and allowed midwives and nurses to insert IUDs (the favorite contraceptive of the Ministry of Health). Nevertheless, the general conclusion from an examination of all the information is that birth control has received rather limited assistance from the state up to the present time -- though more so in the 1980s than in the 1960s and 1970s.

To learn how the population by its own actions, rather than by acts of the state, was able to realize the low fertility levels that it perceived to be necessary for a good life in Turkey, it seems to be necessary to study culture, and to make use of ethnographic methods. For this paper, we have not done this. However, once one starts to listen from the bottom rather than to start from the top, some interesting and rather unexpected questions are raised. (footnote: A project is in progress at Bogazici using this approach, but it will be some time before results are available. I want to thank Zeynep Angin for her contributions, in particular to this last section, though I take responsibility for the final formulations.)

First, does the predominance of male methods of birth control signify that men are in control of the fertility decline in Turkey and that their spouses' roles are diminished by these practices? The answer, we believe, will be a qualified negative, since it is probably the social construction of men's and women's wishes for children and of the means to manage their arrival that is important. A female role as well as a male one are most likely involved, because the actions of the two partners are a matter of negotiation between them. The balance of power and forms of cooperation between the two genders would be a valuable subject to study.

Second, we wonder why the medical establishment has been unable to persuade more couples to use medical methods of control to reduce their fertility. One answer might be that this establishment really does not care; it is only the family planners who are active on behalf of contraception, constructing this to be female contraception. The culture of the Turkish medical establishment itself ought to be studied in order to find answers. (footnote: The Japanese reliance on abortion is not really a "cultural preference" but a consequence of the internal machinations of the medical establishment that kept alternatives out of reach and while physicians pursued the medical business of doing the abortions. See Coleman (1983) cited by Greenhalgh (1990: 19).)

Our preliminary ethnographic work shows that some men and women are themselves cautious about using medical interventions that they think might be harmful to the female body. The aversion to medical contraception is more complex and nuanced than that. Some female methods would place more responsibility and control over births in the hands of women. There is a great deal to investigate here, and these are only hints.

Controlling family size, protecting one's health, regulating one's reproduction, and engaging in sexual activity, emerge as subjects that should not be examined on a purely bio-technological or demographic level. The people who do these things construct their own images of what the goals should be, what signifies good and bad behavior, how men and women should relate and negotiate with each other, and how they should understand the having or not having of children. Until we look more deeply into their social constructions of reproduction, I doubt whether we can understand how and for whom low fertility became the objective, and how it happened to be achieved without very much recourse to modern technology.

(return to contents)

References

Barlow, Robin. 1994. "Population growth and economic growth: Some more correlations." Population and Development Review 20(1): 133-165.

Behar, Cem. 1980. Türkiye'de Nüfus Planlamasi Politikasinin Nüfussal Etkinligi (1965-1980): Bir Uygulama Denemesi (The demographic effectiveness of population planning policy: a policy trial). Docent's dissertation. Istanbul: Faculty of Administrative Sciences, Bogazici University.

Bulut, Aysen, and Nahid Toubia. 1994. Hastanelerde Gebelik Sonlandirma Hizmetlerinin İşlerliği ve Etkinliği (Efficiency and Functioning of Services for Hospital Termination of Pregnancy), Istanbul: Istanbul Universitesi Cocuk Sağliği Enstitüsü.

Coale, Ansley J. 1969. "The decline of fertility in Europe from the French Revolution to World War II," in S.J. Berman, Leslie Corsa, and Ronald Freedman (eds.), Fertility and Family Planning: A World View. Ann Arbor: University of Michigan Press, p 3-24.

Coleman, Samuel. 1993. Family Planning in Japanese Society: Traditional Birth Control in a Modern Urban Culture. Princeton: Princeton University Press.

Dowrick, Steve. 1992. "Technological catch up and diverging incomes: Patterns of economic growth 1960-88," Economic Journal 102: 600-610.

Finkle, Jason L., and C. Alison McIntosh (eds). 1994. The New Politics of Population: Conflict and Consensus in Family Planning. Supplement to Vol 20, 1994 of Population and Development Review.

Greenhalgh, Susan. 1990. "Toward a Political Economy of Fertility: Anthropological Contributions," Research Division, Working Papers No. 12. New York: The Population Council.

Greenhalgh, Susan. 1994. "Anthropological Contributions to Fertility Theory," Working Papers: Research Division, No. 64. New York: The Population Council. Forthcoming in S. Greenhalgh (ed.) Situating Fertility: Anthropology and Demographic Inquiry. Cambridge: Cambridge University Press, 1995.

Hacettepe Institute of Population Studies. 1989. 1988 Turkish Population and Health Survey. Ankara: Hacettepe University.

Hacettepe Institute of Population Studies. 1994. Turkish Demographic and Health Survey, 1993. Co-authors: Ministry of Health, General Directorate of Mother and Child Health and Family Planning, and Macro International, Demographic and Health Surveys. Ankara: Ministry of Health.

Hodgson, Dennis. 1983. "Demography as social science and policy science," Population and Development Review, 9(1): 1-34.

Knodel, John, and Etienne van de Walle. 1979. "Lessons from the past: policy implications of historical fertility studies," Population and Development Review 8(2):217-245.

McNicoll, Geoffrey. 1994. "Institutional Analysis of Fertility," Research Division Working Papers No. 62. New York: The Population Council.

McNicoll, Geoffrey. 1980. "Institutional determinants of fertility change," Population and Development Review 6:441-462.

Ozbay, Ferhunde, and Frederic C. Shorter. 1970. "Turkey: Changes in Birth Control Practice, 1963 to 1968," Studies in Family Planning, 51:1-23.

Santow, Gigi. 1993. "Coitus Interruptus in the Twentieth Century," Population and Development Review, 19(4): 767-792.

Shorter, Frederic C. 1985. "The Population of Turkey After the War of Independence," International Journal of Middle East Studies, 17:417-441.

Shorter, Frederic C., and Miroslav Macura. 1982. Trends in Fertility and Mortality in Turkey 1935-1975, Washington: Committee on Population and Demography, U.S. National Academy of Sciences, National Academy Press.

State Institute of Statistics, Prime Ministry. 1991. 1989 Turkish Demographic Survey. Publication No. 1483. Ankara: SIS.

State Institute of Statistics, Prime Ministry. 1995. The Population of Turkey, 1923-1994: Demographic Structure and Development, with Projections to the Mid-21st Century. Ankara: SIS.

Tarzi, Pakize I. 1992. Anilar (Memoirs), Istanbul: Dizgi Baski.

Watkins, Susan C. 1986. "Conclusions," in A.J. Coale and S.C. Watkins (eds.), The Decline of Fertility in Europe. Princeton: Princeton University Press, p. 420-449.

White, Jenny B. 1994. Money Makes Us Relatives: Women's Labor in Urban Turkey. University of Texas Press, Austin, Texas.

(return to contents)