Fertility Policy Fertility policy is much more complicated, ethically speaking, than mortality policy. Everyone knows that promoting mortality is unethical. Thus mortality policy is concerned only with saving and extending life. When it comes to fertility policy, values differ. Fertility has been both discouraged and encouraged by governments at different times and places, not usually for its own sake, but to accomplish population growth or shrinkage. For example, Quebec promoted fertility between 1988 and 1997, with the aim of keeping francophone culture alive in North America. Many people believe that fertility is a deeply personal individual freedom that government has no business tampering with. Many believe that childbearing is a divine imperative that should not be impeded. We are generally content to allow a government its incentives and advertising regarding fertility, as long as government does not violate our human rights by coercing us, deceiving us, or manipulating us to do something we do not want to do but are too poor to resist doing. The question of coercion comes up in the abortion debate, where the contested right to life of a fetus/unborn child clashes with the contested right of a woman to abort a fetus/unborn child. The issue of which of these rights are valid and, if both are valid, which right prevails, is an important one. However, at various times and places, abortion legislation has been enacted not to answer this question but to achieve a target level of fertility in the population. The Ceausescu regime in Romania (1965-1989) outlawed abortion to achieve a higher birth rate. Meanwhile, Singapore legalized abortion in 1969 for the express purpose of reducing births. Forced abortion and sterilization have occurred in China since the 1980s for the same reason. Another example of coercion is the forced sterilization of mentally ill and mentally retarded people in North America and northwestern Europe between the World Wars. A more recent example is the forced sterilization of poor men – particularly Muslim men - in India in the mid 1970s. Those episodes remind us that many times, fertility policy – like immigration policy – is directed at particular groups of people: ethnic groups, religious groups, or income classes. It is those sub-populations that are targeted for growth or shrinkage. Thus the question of discrimination is another issue complicating fertility policy. Eugenics Nazi Germany went furthest in elaborating an ideology of genetic superiority. The fallout – millions killed on the basis of their race, politics, religion, color, intelligence, sexual orientation etc. – served as a wake-up call to the would-be civilized world. It was not only the Germans, but citizens of many nations who embraced eugenics, including US President Theodore Roosevelt, Planned Parenthood founder Margaret Sanger, and Irving Fisher (celebrated economist). In Canada, the eugenics movement had most influen ...

1. Fertility Policy
Fertility policy is much more complicated, ethically speaking,
than mortality policy. Everyone knows that promoting
mortality is unethical. Thus mortality policy is concerned only
with saving and extending life.
When it comes to fertility policy, values differ. Fertility has
been both discouraged and encouraged by governments at
different times and places, not usually for its own sake, but to
accomplish population growth or shrinkage. For example,
Quebec promoted fertility between 1988 and 1997, with the aim
of keeping francophone culture alive in North America.
Many people believe that fertility is a deeply personal
individual freedom that government has no business tampering
with. Many believe that childbearing is a divine imperative that
should not be impeded.
We are generally content to allow a government its incentives
and advertising regarding fertility, as long as government does
not violate our human rights by coercing us, deceiving us, or
manipulating us to do something we do not want to do but are
too poor to resist doing.
The question of coercion comes up in the abortion debate,
where the contested right to life of a fetus/unborn child clashes
with the contested right of a woman to abort a fetus/unborn
child. The issue of which of these rights are valid and, if both
are valid, which right prevails, is an important one. However,
at various times and places, abortion legislation has been
enacted not to answer this question but to achieve a target level
of fertility in the population. The Ceausescu regime in Romania
(1965-1989) outlawed abortion to achieve a higher birth rate.
Meanwhile, Singapore legalized abortion in 1969 for the
express purpose of reducing births. Forced abortion and

2. sterilization have occurred in China since the 1980s for the
same reason.
Another example of coercion is the forced sterilization of
mentally ill and mentally retarded people in North America and
northwestern Europe between the World Wars. A more recent
example is the forced sterilization of poor men – particularly
Muslim men - in India in the mid 1970s.
Those episodes remind us that many times, fertility policy – like
immigration policy – is directed at particular groups of people:
ethnic groups, religious groups, or income classes. It is those
sub-populations that are targeted for growth or shrinkage. Thus
the question of discrimination is another issue complicating
fertility policy.
Eugenics
Nazi Germany went furthest in elaborating an ideology of
genetic superiority. The fallout – millions killed on the basis of
their race, politics, religion, color, intelligence, sexual
orientation etc. – served as a wake-up call to the would-be
civilized world.
It was not only the Germans, but citizens of many nations who
embraced eugenics, including US President Theodore Roosevelt,
Planned Parenthood founder Margaret Sanger, and Irving Fisher
(celebrated economist). In Canada, the eugenics movement had
most influence in Alberta, where a Eugenics Board, with the
authority to sterilize people deemed defective, operated between
1928 and 1972.
The damage done by such policies, the horror of Nazi camps,
the exploitation, imprisonment and killing of citizens by
Communist dictators, and books such as “Brave New World”
(Huxley, 1932) have warned us that governments’ social
planning can completely override compassion and respect for
human rights.
Thus in today’s world, most governments have given up
supervising reproductive selection. However, technology –

3. combined with liberal abortion laws – is giving prospective
parents the opportunity to themselves screen their offspring for
unwanted characteristics.
This is a lot safer than government screening: there is a
diversity of parents who will welcome children like themselves,
preserving diversity. However, it is not clear that girl children,
and children with congenital disabilities, will be as likely to
make the cut as boys and healthy children. In the future it may
become easier to select for all kinds of apparent abilities and
advantages.
Economics cannot be relied upon to dissuade parents from this
course of action, especially if children with unwanted
characteristics cost more to raise and governments are not
willing to share the burden. However, economics can remind us
that diversity is a source of strength, providing fresh ideas and
approaches as well as opportunities for specialization. Our
society's strength, our government's strength, our economy's
strength, lies in diversity, cooperation, and competition, rather
than in conformity, coercion, and cronyism.
Understanding that we mean no coercion and no discrimination,
let us discuss how fertility policy might be implemented
successfully.
Five Principles of Policy Design
#1 No discrimination or coercion. As discussed above.
#2 Question the policy. Identify the ultimate goal or the root
cause of the problem you are trying to address. Is the proposed
policy the most direct way to achieve your goal/fight the
problem? For example, if you are embarking on a program to
reduce births, there is probably a deeper goal, such as poverty
reduction. A policy to encourage births might really be about
increasing the labour supply. There may be more direct and

4. faster-acting ways to reduce poverty or increase the labour
supply.
#3 Target the binding constraint
A successful policy addresses the most critical bottleneck, the
most pressing barrier to achieving the goal or reducing the
problem at hand. For example, if the policy is intended to
encourage births, you need to know what is really holding
people back from deciding to have children or expand their
family size. It’s no use offering money to couples to have
children, if they are avoiding children for non-financial reasons.
#4 Target the appropriate margin
In microeconomics we learn that people evaluate things at the
margin. They decide whether or not to study one more hour, not
just whether or not to study at all. They decide whether or not
to have kids, but then they decide whether or not to have one
more, one at a time. If most people already intend to have one
child, you should target people who are at the margin of
deciding for another child or not. Similarly, if your target is a
limit of two children per family, you can implement a policy
that discourages third children.
Poster of Singapore Family Planning and Population Board,
1978.
Another margin that is relevant to fertility is hours worked. Is
the parent deciding whether to join the workforce or whether to
work a few more hours each day? Is the parent already
committed to working full-time no matter what?
#5 Understand who pays the financial cost. Remember that
taxes and subsidies always affect both producers and
consumers, no matter which of them has the tax or subsidy
imposed on them. The least price-sensitive party pays most of
the tax. The least price-sensitive party gains most of the
subsidy.
This means that subsidies intended to encourage fertility may be
ineffective if the supply of houses, childcare spaces, etc. is
inelastic. It means that taxes to discourage fertility will not be
effective if fertility is price-insensitive; instead, those having

5. children will pay the tax and their ability to look after the
children will be compromised.
In class we will evaluate various pro-natalist and anti-natalist
policies including cash incentives, subsidized daycare, cash-for-
care (Norway), paid parental leave, subsidized birth control, and
the general education of girls. We also learn about China’s One
Child Policy.
Chinese Population Policy
China’s first official family planning programs were in place in
the late 50s in some large cities, encouraging couples to plan
the number of children and choose fewer. China’s formal one-
child policy, begun in 1979, has been possibly the most focused
and wide-ranging birth control program ever. The program was
launched with Chairman Deng’s announcement of a zero
population growth (ZPG) target for 2000. To this end, births
were to be limited to one per couple, with some exceptions.
Program Details
The program stipulates of a maximum number of children
permitted, depending on region and ethnicity. Each child
requires a birth permit.
Table 34-1. One-Child Policy Details.
Group
Regulation
urban residents
one child
most rural residents
two children if the first child is female or handicapped; or two
children four years apart in age
minorities in minority autonomous regions
two or three children

6. rural Tibetans
any number of children
The program offers economic incentives for compliance, which
have included urban one-child families receiving a monthly
allowance until the child is 14 years old, plus preferential
housing, school admissions, and pensions. Rural one-child
families received extra work points until the child was 14 years
old, and the same size grain ration and size of plot as 2- child
families.
At first, the program was administered centrally, relying on
propaganda and yearly “shock drives” –which included forced
sterilizations and abortions - to achieve local targets. This led
to fierce confrontations. At the same time, market reform was
occurring and making peasants less dependent on government
subsidies. Peasants stood to gain personally from additional
sons to work the land. Consequently, in the mid-80s the targets
for rural couples were relaxed. It took until 2001, however, for
coercion to be officially prohibited. This came as a result of
domestic clashes, international pressure, better demographic
data, and positive results from pilot projects which concentrated
on providing information and health care, say Zhao and Guo
(2007). The new policy also prohibits sex-specific abortion and
discrimination against female children. Compliance is
imperfect.
Results of the Program for Fertility and Population Growth
Chairman Deng’s original goal, ZPG by 2000, was not achieved.
The growth rate in 2000 was 1.07%, not a whole lot less than
the 1981 level of 1.4%. By 2009 the population growth rate had
fallen even more to 0.61% (Canada had 0.82%). It is estimated
that the Chinese population is now significantly smaller that it
would have been without Deng's policy, by hundreds of millions
of people. To calculate what population would have been
without the policy we would need to run a Leslie matrix over

7. the length of time the policy has been in place, using the
original fertility rates, but adjusting mortality rates as they
changed over time. (However, even without the one-child
policy, fertility rates might have dropped with economic
development.) Representatives of the Chinese government,
which claims that 400 million deaths were averted over 30 years
(Lifesitenews.com, 2006), have suggested that China has
already made its contribution to fighting climate change.
China's TFR is about 1.9, down from 2.7 in 1980, and well
below replacement TFR of 2.1 children per female. This
decrease was critical because it helped defuse the population
momentum that existed due to a baby boom that took place in
the late 60s. The number of people of childbearing age will not
decline until 2015.
What would happen if the one child policy were abandoned? In
fact the policy is becoming less rigid. Currently, urban Chinese
couples are permitted a second child, if each person in the
couples was himself or herself the only child in his/her family.
As the Chinese population continues to age, and as its people
achieve new political freedoms, more children will be
permitted. New freedom will also allow the parents to purse
new and varied careers. It will be interesting to see to what
extent the government’s one-child program has been taken to
heart by the Chinese people.
Missing Females
There have been many side effects of the drive to lower
fertility.
We have already mentioned that forced abortions and
sterilizations have taken place, and we can imagine the scars
that are left behind.
Another serious problem linked to the program is girl-specific
abortion, infanticide, and neglect. Though most parents in
China treasure their girl children, cultural values and economic
pressures lead some parents to prefer boys and to do away with

8. girl children in hopes of being able to have a son instead. In
some provinces, typically those having large rural non-minority
populations, the sex ratio at birth may be as high as 119
compared to 105 in other parts of China or 107 in Tibet.
Selection for boys may not be entirely the fault of the one-child
policy. Other nations, such as South Korea, where son
preference is declining from its 1990 high, and northewestern
India, where it is stronger than ever
, also have skewed sex ratios
Table 34-2. Selected Sex Ratios
Sex ratios:
Age 0-4, 1982
Age 0-4, 1995
Age 0-4, 2005
Beijing, China
107.3
113.5
112
Anhui Province, China
110
125.1
136.4
Xinjiang Province, China
103.7
101.8
1105.5
Sex ratio:
Age 0-6, 2001
Age 0-6, 2011

9. India
107.9
109.4
Punjab Province, India
125.3
118.2
Dadra and Nagar Haveli Provinces, India
102.1
108.2
Sex ratios:
At birth, 1981
At birth, 1989
At birth, 1992
At birth, 2001
South Korea
104
112
114
108
Sources: Das Gupta et al. (2009), Census of India (2011),
Hesketh and Zhu (2006).
Das Gupta (December 2009) argues that Korea, China, and
northwestern India, places where son preference has manifested
itself especially strongly, not only have been patrilineal (only
men inherit), but moreover have had traditional political
systems which are very much organized around male ancestry
Ancestor worship helped reinforce notions of loyalty, order, and
political hierarchy. In rural areas these values still hold sway
and a man’s identity, social status, and access to resources is
determined by his position in a clan. For example, the oldest

10. son of an oldest son is in a favoured position. A woman’s
identity is determined by her husband. Women born into the
clan are required to leave and marry men of other clans. They
leave their land and forego any inheritance other than what is
given to them as part of the marriage settlement.
Traditionally, Asian women live with their in-laws once
married, so it is their brothers who look after their aging
parents. In the School of Policy Studies at Queen’s, Wei Li
Ding studies rural access to credit in China. She finds that
families with sons have an easier time getting loans. One
reason is that sons are more likely to be able to earn money and
share that money with parents.
Living with in-laws, a women is dependent on them for
protection, sustenance, and approval. The husband’s parents are
likely to influence her and her husband’s fertility decisions.
In India, the advantage of having a son is heightened by the
necessity of paying a dowry to the groom’s family when a
daughter gets married. As one advertisement for a fetal-gender
test kit put it, "Spend 500 rupees now to save 500,000 rupees
later."
Earlier we described the research of Jiang, Feldman, and Jin
(2005), who estimated the number of Chinese females missing
over the last century. Jiang et al. conclude that 35 million
Chinese females were lost over the century, about 4.65 percent
of all females who were expected to be born. The number of
missing females steadily increased during the years of the One-
Child policy. See their Figure 1 on the next page.
The number of missing females may be exaggerated if girls and
women are under-reported.

11. For Asia as a whole, it is estimated that 163 million females
that should be present are not. (Hvistendahl, 2011). The
Economist
predicts that by 2025, China will have only 80.3 million woen
in their twenties compared to 96.5 million men in their twenties
(a sex ratio of 1.2). One might think that increasing scarcity of
females will lead to increasing brideprice (the traditional
Chinese norm) at marriage and an increasing appreciation of the
role of women, with wives being treated better. Unfortunately,
lacking individual rights and freedoms, many women will be at
higher risk for being kidnapped, pimped, or forced into
monogamous or plural marriage.
There are also negative consequences for men. Many will
remain involuntarily single. Single men generally have poorer
health and earlier death than married men. They may have to
spend more resources or take bigger risks to attract a bride.
They may have to migrate to find a partner, or settle for one
who is less compatible. For society as a whole, tension and
unrest may increase. Fertility will be lower than otherwise
because of the absence of so many women.
Following page: Figure 34-2. Females missing from China, as
percent of population.
Source: Jiang, Feldman, and Jin (2005)
Regarding Figure 34-2, the following dates are of interest:
1910- slavery abolished
1911- Sun Yat Sen leads revolution against Qing Dynasty
1916+ warlord era
1931-1945 Japanese occupation
1949 Communist Party is established as the government
1957-58 Great Leap Forward and famine

12. 1966-1976 Cultural Revolution
1976 Death of Chairman Mao
1979 One Child Policy instituted
continued
Other consequences of the One Child Policy:
Population composition: Aging population
As birth cohorts fall in size, the population ages. Although the
overall dependency ratio in China fell between 1982 and 2000,
the aged dependency ratio rose, though at 0.11 it is still lower
than Canada’s aged dependency ratio of 0.21. Yet China is
experiencing a level of aged dependency usually associated with
more economically developed nations.
China’s extensive social welfare system, concentrated in the
cities, is being strained. Health care in rural areas will be a
challenge.
Population composition: fewer children from urban families.
Unless rural areas receive the same educational opportunities as
urban, the proportion of the population which is educated may
fall if rural families have more children than urban families.
Population composition: little emperors. Some have worried
that children with no siblings will be pampered and less socially
conscious. On the other hand there may be benefits that come
from being raised in an adult-intensive environment and
receiving relatively more adult attention. These concerns are
beyond the scope of our course!
Population deceleration: reduced rate of capital shallowing.
China is currently a low-wage country with a low capital:labour

13. ratio; there is also a housing shortage, an education shortage,
and problems of environmental degradation. Yet consider how
much worse these problems could have been had fertility not
been discouraged. Though the workforce now is smaller than it
might have been otherwise, machines, land, and education per
person are higher.
� There are some signs of hope at the sub-national level. See
Punjab Province in Table 34-2.
� As reported in "Land of the rising son", Globe and Mail,
Sept. 12, 2009.
� Some dates to consider: 1911: Revolution against Qing
Dynasty begins. 1916: Warlord era begins. 1931-1945:
Japanese Occupation. 1949: Communist Party of China takes
control. 1957/8: Great Leap Forward leads to mass deaths.
1966-1976: Cultural Revolution. 1976: Deng's economic
reforms begin, followed by One Child Policy in 1979.
� “A tale of three islands,” October 22, 2011.
Measuring Fertility
The general fertility rate is measured differently from the
birth rate, with the denominator showing not mid-year
population, but the mid-year population of females of
childbearing age.
General fertility rate = (# births/midyear population of females
aged 15-49) x 1000

14. Age-specific fertility rates (ASFR) give even more precision as
to the age. For example, the (age-specific) fertility rate for
females 15-19 years old ≡ number of live births to 15-19 year
olds during the year / mid-year population of female 15-19 year
olds, all multiplied by 1000.
Though in general, age-specific fertility rates have been
dropping in Canada, an exception is the fertility rate for women
over 30. Those rates have been rising since the late 70s. In
fact, the average age of mother has been rising since the late
70s (see Figure 28-4).
Figure 28-1. Age-Specific Fertility Rates, Canada
Source for data: Statistics Canada, “Fertility rate by age group,
Canada, 1926-2008” in Fertility: Overview, 2008,
http://www.statcan.gc.ca/pub/91-209-x/2011001/article/11513-
eng.htm downloaded October 31, 2013.
Once a woman, or a group of women born the same year, is no
longer of child-bearing age, we can record how many children
were born to that woman or cohort. We can compute the
“completed fertility rate (CFR)” which is children per woman,
for that group of women.
CFR = # children born to a group of women/ number of women
in the group
For example, for women born in 1946, the completed fertility
rate was 2.1, i.e. 2.1 children per woman.
But that is looking into the past. To get an idea of how many
children today’s women will have, demographers compute a
hypothetical statistic called the total fertility rate (TFR). The

15. total fertility rate is the number of children which would be
born to the average woman IF the average woman experiences
today’s age-specific fertility rates at each age of her life. This
is not completely realistic. In 2010, 25-year old women will
behave as predicted by the 2010 fertility rate for 25-year old
women. But by 2015, when the women are now 30, we cannot
expect their fertility to match the fertility of 30-year-old women
in 2010. They will make their own decisions.
TFR = ∑ ASFR over all age groups x 5 / 1000
where ASFR is age-specific fertility rates
“5” represents the 5 years a woman spends in the typcial age
group. Age groups are usually 5 years in length e.g. 15-19, 20-
24, 25-29 etc.
Why do we divide by 1000? Well, the ASFR gives you the
number of children per 1000 women of that age, say 130
children. Now one woman cannot have 130 children. Only
1000 women can. So we divide by 1000 to get the children per
woman.
Figure 28-2. Total Fertility Rate, Canada
Source: WDI Online, World Bank Group
As you can see in the Figure above, Canada's TFR was about 1.5
in 2005. It was slightly higher, at 1.58, in 2010.
If we were to count only the female babies in our ASFR, and
then compute TFR, we would have the Gross Reproduction
Rate, or number of female babies per woman. If we went a step
further and multiplied the female baby ASFR by the probability
of a female baby living to its mother’s age group, we would
have the Net Reproduction Rate. A population is self-

16. sustaining if its NRR is greater than or equal to one.
Instead of using NRR, we usually compute TFR and
consider a TFR of 2.1 to be sufficient for a population to be
self-sustaining. A TFR=2.1 is considered “the replacement
rate” or “replacement fertility”. The last Canadian cohort to
achieve a CFR of 2.1 was the women born in 1946. Subsequent
cohorts of women have had fewer than 2.1 children per woman.
Tempo-adjusted TFR
Because TFR exaggerates fertility decline when the age of the
mother is increasing, demographers have developed a tempo-
adjusted TFR.
Adjusted TFR (t) = TFR (t)/ (1-r(t) )
where r(t) measures the influence of postponing fertility.
r(t) =
average age of woman giving birth (t+1) – average age of
woman giving birth (t-1)
Demographers refine this calculation by first computing TFR
for one kind of child: eldest, second, third, etc. Such a TFR is
called a birth-order specific TFR.
Figure 28-3. Fertility trends in the Czech Republic, showing
tempo-adjusted TFP.
Source: Philipov and Sobotka (2006)
In Figure 28-3 we see that, when adjusted for the increasing age
of mothers, Czech fertility rates are higher than they originally
appeared to be.

17. Figure 28-4. Average age of mother at childbirth, Canada,
various years.
Source: Human Resources and Skills Development Canada,
2011.
Determinants of Fertility
Fertility, or how many children per female are born, depends on
three things: opportunities for intentional or unintended
procreation; intentions; and ability to carry out those intentions.
The decision maker is usually a heterosexual couple relying on
their own powers of procreation. In some cultures, the parents
of the husband traditionally have influenced a couple’s fertility
decisions. Modern western couples have great autonomy in
fertility. Greater personal freedom, greater social tolerance of
unusual families, and medical technology have united to make it
possible for infertile couples, homosexual couples, and singles
to become parents.
Opportunities
Procreation, both intentional and unintentional, requires one
man and one woman. There is now the possibility of women
using sperm banks to conceive, or a couple using a surrogate
mother to carry a child to term. Traditional factors governing
the union of men and women include the degree of social

18. isolation of men or women; sexual activity rates; the sex ratio;
the usual age at marriage or cohabitation; types of marriage
(e.g.polygamy vs. monogamy); absence of spouse; likelihood of
bereavement, separation, or divorce; and time between unions.
Social and religious norms, income, geography, and political
crises influence these things. Generally speaking, prosperity,
peace, and secularization mean greater opportunities for
coupling and conception.
Intentions
For the decision maker(s), the target number of children
depends on personal preferences, social and religious norms,
and economic considerations. It also depends on a person's
experience of childhood and of raising any previous children.
When infant and child mortality is high, extra children may be
born in order to achieve the target number.
Extra children are sometimes born to achieve a target
number of children with particular characteristics. In the past,
one had to “keep trying” until one had the requisite number of
boys, girls, healthy children etc. Now genetic testing in utero is
making selection easier, and reducing overall fertility. It is
ironic that, just when we are treating peoploe with disabilities
with dignity,and making it easier for them to take their place in
society, we have the means, and often the will, to make sure
they are never born.
Even before modern birth control methods were available,
when contraception was limited to herbal teas, abstinence, and
withdrawal, major declines in fertility took place in
industrializing societies as the demographic transition
progressed. For example, births per year per married woman in
France fell from 0.775 in 1740 to 0.410 in 1891 to 0.273 in
1931 (Wrigley, 1985). Scholars collaborating on the European
Fertility History Project in the 1960s produced nine books on

19. the demographic transition in Europe. They concluded that
secularization, more than anything else, was associated with
falling fertility rates. Thus, fertility fell in European provinces
where infant mortality was still high, and where income per
person had not yet risen appreciably, if those provinces were
integrated with more secularized provinces sharing the same
language and culture.
Secularization of a culture occurs as society organizes
itself along non-religious lines. The government, the courts,
schools, and the like adopt a neutral religious stance.
Secularization encourages education, personal development, and
decision making without reference to religious authority. It
encourages individualism, sometimes at the expense of the
community.
Ability to achieve intentions
Although Europeans were able to reduce fertility without
modern birth control methods, it is no doubt easier to limit
family size today.
Birth control includes contraceptive methods, sterilization of
the male or female, and abortion. Abortion continues to arouse
serious concern, especially for religious thinkers. The use of
abortion to select for boys troubles secular thinkers too.
Infanticide, neglect, and abandonment may be forms of delayed
birth control.
Some parents, while wishing to reduce family size, reject
some or all methods of birth control on moral or medical
grounds. Others are ignorant of birth control methods or find
them inaccessible. Others simply procrastinate or lack the
discipline to use any method of birth control consistently.
The “fertility trap” refers to the phenomenon of delaying
child-bearing - perhaps because of a shock such as the collapse

20. of Communism and its parental support programs - until the
parents’ fertility has declined or there are simply too few years
to give birth to the number of children originally desired. The
new, smaller-than-intended families foster new social norms
about family size, parental age, and parental career development
leading to continuation of the small-family pattern.
Indeed, it is infertility – difficulty conceiving a child – that is
perhaps the greatest barrier today to achieving desired family
size. Parents must be healthy and physiologically able to
conceive a child and carry it to term. Breastfeeding,
malnutrition, disease, excessive exercise and stress can interfere
with conception and pregnancy. Adoption and fostering are
alternative ways to building a family. Adoption and fostering
are not part of fertility, but they do contribute to a healthier,
more emotionally resilient population.
Economic prosperity makes it easier for people to manage their
health and to access birth control. It therefore enhances the
ability of parents to achieve a desired family size. However,
recent history suggests that intended family size may shrinks
with secularization and economic growth.
Economic Influences on Desired Family Size
What benefit and cost considerations affect the number of
children a family chooses to have?
Children providing material benefits
First we might ask whether children provide any material
benefits to parents. In various times and places children may
have been perceived as a net material benefit to parents.
Children can work for the family and children can care for their
parents when parents are no longer able to work themselves.

21. In her book on British children brought to Canada, Joy Parr
notes that children under the age of 14 could not be boarded out
at a profit. Becker (1960) mentioned a finding that male slaves
were a net expense to American slave owners until those slaves
were about 18 years of age. However, children of about 8 years
of age and older have worked in sweatshops and on plantations
as long as poverty has existed. Today their tiny fingers weave
carpets in Afghanistan and their tiny bodies wriggle through
diamond tunnels in Tanzania. At some point they may yield a
net profit to their parents.
Whether or not money can be recouped from children before
they reach adulthood, adult children can represent a safety net
for parents. In societies where healthcare is unsubsidized,
insurance is unaffordable, and pensions are inadequate, parents
may look to children to meet the needs of their aging.
However, in such societies, the capital:labour ratio is likely to
be low, and the returns to additional capital greater than the
returns to additional labor. If therefore the government could
create reliable capital markets, safe vehicles for savings,
insurance programs, and pension plans, citizens would dare
divert money away from children and toward investments that
would provide greater material benefit for themselves and their
society.
In this day and age - urban, prosperous, and human rights-
oriented - it seems more natural to look at children providing an
emotional rather than a financial benefit to their parents.
Children providing psychic benefits to their parents
With children providing “utility” we could use the economic
framework of utility maximization within the limits of a budget
and a 24 hour day. The decision maker’s constrained utility
maximization results in a desired number of children as well as
a desired amount of alternative goods and leisure opportunities.

22. The key determinants of the demand for children are income,
the cost of children, the cost of substitute goods, and the cost of
complementary goods. This is a crass, incomplete, but
suggestive approach to understanding the fertility decision.
This approach assumes that children are actually “goods”, i.e.,
that more children are preferred to fewer. Is this indeed the
case?
The following figure uses a large 2007 survey of american
heterosexual couples where the female is under 55 years of age.
Figure 30-1. Average number of children, by income rank of
families.
Data source: Panel Study of Income Dynamics (PSID),
2009.[footnoteRef:1]
[1: Panel Study of Income Dynamics public use dataset.
Produced and distributed by the University of Michigan with
primary funding from the National Science Foundation, the
National Institute of Aging, and the National Institute of Child
Health and Human Development. Ann Arbor, Michigan, 2009.
]
A brief glance at Figure 30-1 above suggests that, inasmuch as
children are “goods” which enter a parent’s utility function, it is
not clear they are normal goods, i.e. goods which are purchased
in greater number as income rises. Roughly speaking, lower
income families have more children. We also observe that lower
income countries have more children than higher income
countries.
Willis (1973) developed a detailed model of parental choice. In
Willis’ model, each parental unit maximizes utility which

23. depends on the N, the number of children; Q, the level of
childhood “quality” for each child which costs time and money;
and S, an alternative activity such as skiing. Parents choose
between spending on skiing (S) or on child services (N*Q).
Each parental unit also faces constraints. Parents have wage
income, and a limited amount of time which for one of the
parents – traditionally, the woman - is allocated between work
and childcare. Her wage depends on her initial level of skill and
her experience in the workforce. Time spent with children
means no wage now and a lower wage in the future.
Note that Willis' model is not appropriate for a less-
economically-developed nation. Children provide no labour.
And there is no consideration of what happens to parents in
their old age when they cannot work and when perhaps there is
no government support for them.
In Willis' model, if there is an increase in endowment income,
and child services is a normal good, the parents will want more
child services (NQ). However, we do not know whether more
child services means more children, or more money spent on
each child. The more they spend on existing children, the more
expensive additional children will be, if parents want to treat all
children equally. If parents get richer, they might likely choose
to increase Q without increasing N. We can thus have
increases in income translating into fewer children without
children being “inferior goods”.
It is possible that for some people, children are indeed inferior
goods, to be foregone or abandoned if more exciting
opportunities come along. Another way to explain falling
fertility rates for higher income people and nations is that, for
them, children are not needed for material benefits, only for
psychic ones.

24. According to Willis’ model, the following things make it likely
fewer children will be desired:
-a decrease in the cost of S, where S is a substitute for
childservices
-an increase in the time or money cost of NQ
- a high desired Q for each child.
-any decrease in the father’s lifetime earnings.
-if child-rearing is more time-consuming than S, but less
expensive than S, an increase in the mother’s wage.
Did you notice the difference between the effects of the father’s
and mother's earnings? The father’s earnings were assumed to
be independent of the time spent raising the children. Since the
father’s time was not the subject of optimal allocation between
work and children, it could be treated as manna from heaven, a
source of “endowment income” that allows more of everything
to be acquired.
The mother’s earnings, however, involve a tradeoff: less time
spent at home. They have an opportunity cost: time spent with
children.
Substitution and Income Effects
Recall from microeconomic theory that when the price of
something changes, there is an income effect and a substitution
effect. Consider an example. When the price of apples rises,
you switch to cheaper fruits. This is the substitution effect.
From the point of view of a consumer, you feel poorer because
your purchasing power has fallen. You buy fewer apples and
fewer of any goods that are normal. This is the income effect,
which, for consumers, works in the same direction as the
substitution effect.
For a vendor, however, the substitution and income effects
work in opposite directions. A vendor of apples, when
snacking, will buy cheaper fruits when the price of apples rises.

25. The substitution effect causes him to buy fewer apples.
However, a rising price of apples will make the vendor richer
and he or she will buy more apples and more of any other
normal goods. The income effect causes a seller of apples to
buy more apples, not fewer.
When a potentially care-giving parent is offered a
wage increase, there are two effects. The substitution effect
tells the parent to work more now that the price (opportunity
cost) of her time has risen, and childcare at home is now more
expensive. However, the income effect makes her, as a vendor
of time, feel richer, and able to afford more time spent with
kids. If her partner’s wage rises, again she experiences the
income effect telling her she can afford more time with the kids.
At low wages, the income effect is probably less powerful
than the substitution effect. So a wage increase leads the parent
to work more/spend less time with children. At high wages the
income effect may be larger than the substitution effect, so that
the parent decides to spend more time at home or have more
children.
T. Paul Schulz, a famous development economist, wrote that,
“There is an inverse association between income per adult and
fertility among countries, and across households this inverse
association is also often observed. Many studies find fertility is
lower among better educated women [implying that the
substitution effect outweighs the income effect] and is often
higher among women whose families own more land and assets
[a pure income effect].”[footnoteRef:2] [2: Schultz (2005).]
Figure 30-2 shows a declining number of children for wealthier
families, apparently contradicting the positive income effect we
have postulated.

26. Figure 30-2. Average number of children by wealth rank of
families.
Data source: PSID (2009). All heterosexual couples with a
woman less than 55 years of age.
The effect of the wife’s wage on number of children is obscure.
A strong substitution effect is not observed.
Figure 30-3. Average number of children for families ranked by
size of wife’s wage.
Source: PSID (2009). All heterosexual couples with a woman
less than 55 years of age
The Economist (August 8, 2009) describes the research of
Myrskyla et al. (2009), which suggests the income effect
becomes dominant for countries with high socioeconomic
performance. Graphing the total fertility rate against the United
Nations’ Human Development Index (HDI)[footnoteRef:3] for
240 countries, the authors observed that fertility fell as HDI
increased, but only up to a score of about 0.9. For most
countries, Canada and Japan excepted, whose HDIs exceed
0.9,TFR increased as the HDI increased. [3: HDI is an index of
life expectancy at birth, income per person, and education levels

27. achieved]
Abeysinghe (1993) studied Canadian fertility and used
statistical analysis to correlate wages, income, and the number
of children. He concluded that when it comes to female wages,
the substitution effect outweighs the income effect of a wage
increase, and higher females wages mean fewer children. On
the other hand, there is a pure income effect which favours
children: men whose incomes compared favourably to their
parents had more children.
Although age-specific fertility rates and the total fertility rate
fell when female wages rose, the drop in fertility seems to be
temporary. That is, higher wages cause women to postpone
rather than to avoid childbearing. Abeysinghe found that the
female wage rate was not much correlated with the completed
fertility rate. Tempo-adjusted TFR would not be as sensitive to
female wages as TFR itself.
Not everyone who postpones having kids will find the time or
partner, or be fertile enough to have kids later. (Recall the
"fertility trap".) However, many of them will be able to have
their children later on. Inasmuch as that is the case, TFR
underestimates CFR.
Business Cycle Effects
If higher wages are associated with fewer children, at least
temporarily, we would expect fewer births during economic
boom times, and more children during recessions. However this
is not what happens. TFR rises during economic expansions.
Mocan (1990) attributes most of this effect to the fact that
during economic expansions the age at marriage falls and the
divorce rate falls. He believes that fertility itself is slightly
countercyclical due to rising wages as predicted by the

28. substitution effect.
In conclusion
When asked how income affects fertility, we must distinguish
between three different aspects of income:
Table 30-1. Aspects of income which affect fertility.
Wealth
A pure income effect increases the money and time spent on
children. Some studies suggest that the number of children
rises.
Higher wages, better employment possibilities for the
caregiving parent(s)
The substitution effect prevails at lower wages. If wages
become high enough, the income effect may prevail leading to
greater fertility. However, higher wages are usually associated
with fewer children. TFR falls but tempo-adjusted TFR might
not fall as much.
Higher wages during economic boom
The higher wages might lead to fewer children if it were not the
case that divorce rates fall and people get married earlier during
economic booms. Expect an increase in TFR.
Economic development apart from wealth or higher wages
· reduced reliance on children for labour
· reduced reliance on children for old age security
· improved infant and child survival
· improved education and awareness of career options, birth
control
· improved access to birth control and to infertility treatment.
· secularization
Overall, economic development appears to decrease fertility
until high levels of development are reached.
Canada’s Fertility Rate, 1960-2007

29. total fertility rate 1960 1961 1962 1963 1964 1965 1966 1967
1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978
1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
2001 2002 2003 2004 2005 2006 2007 3.8110001087188721
3.753000020980835 3.6809999942779541
3.6070001125335693 3.4560000896453857
3.1150000095367432 2.749000072479248
2.5280001163482666 2.3859999179840088
2.3340001106262207 2.2579998970031738
2.1410000324249268 1.9800000190734863
1.8899999856948853 1.8370000123977661
1.8240000009536743 1.7960000038146973
1.781999945640564 1.7400000095367432
1.7000000476837158 1.690000057220459
1.6799999475479126 1.6499999761581421
1.6699999570846558 1.6799999475479126
1.6799999475479126 1.7699999809265137
1.8300000429153442 1.7000000476837158
1.7100000381469727 1.7000000476837158
1.6390000581741333 1.5920000076293945
1.5499999523162842 1.4900000095367432
1.5199999809265137 1.5299999713897705
1.5299999713897705 1.5399999618530273
1.5900000333786011 1.5900000333786011
The Effect of Urbanization on China’s Fertility
Zhen Guo • Zheng Wu • Christoph M. Schimmele •

30. Shuzhuo Li
Received: 15 September 2011 / Accepted: 24 January 2012 /
Published online: 10 February 2012
� Springer Science+Business Media B.V. 2012
Abstract The relationship between urbanization and fertility
decline is known to
be inverse in developed countries. However, the nature of this
relationship in
developing countries that already have relatively low fertilities
is not well-under-
stood. This study aims to illustrate how much urbanization
contributed to China’s
fertility decline between 1982 and 2008 and forecasts how much
it can contribute to
future reductions in fertility. The study examines changes in the
total fertility rate
(TFR) at both the national and provincial levels, given regional
differences in the
urbanization rate. The results show that changes in rural
fertility behavior accounted
for most of the decline in the national TFR between 1982 and
2008. This finding
suggests that official birth control policies were instrumental in
curbing China’s

31. population growth. However, urbanization was responsible for
about 22% of the
decrease in TFR during this period, and its effect was especially
important during
the latter years (2001–2008). In most provinces, urbanization
associated with a
decline in provincial-level fertility. The forecasts indicate that
urbanization will
become the primary factor behind future declines in national
fertility. Given the
negative effect of urbanization on the TFR, it is possible to
relax the one-child
policy without having adverse implications for population
growth.
Keywords Urbanization � Fertility � China
Z. Guo
School of Management, Xi’an Jiaotong University, Xi’an, China
Z. Wu (&) � C. M. Schimmele
Department of Sociology, University of Victoria, 3800 Finnerty
Road,
Victoria, BC V8W 3P5, Canada
e-mail: [email protected]
S. Li

32. Institute for Population and Development Studies, School of
Public Policy and Administration,
Xi’an Jiaotong University, Xi’an, China
123
Popul Res Policy Rev (2012) 31:417–434
DOI 10.1007/s11113-012-9230-0
Introduction
China’s total fertility rate (TFR) declined from 2.78 in 1978 to
1.47 in 2008 (National
Bureau of Statistics of China 2009). The TFR decreased to sub-
replacement levels
sometime during the early 1990s (Feeney and Yuan 1994). The
pace of this decline is
remarkable considering that the TFR was over 5.0 until the
1970s (Gu 2007). This
change has been attributed primarily to the Chinese
government’s efforts to curb
population growth, such as the one-child policy (Feeney and
Wang 1993). These
birth control interventions certainly set China’s demographic
transition apart from

33. other transitions to low-fertility. However, whether birth
planning policies are the
primary reason for China’s low fertility is not uncontested (Cai
2010). At least, it
appears that, similar to the demographic transitions in Western
countries, socioeco-
nomic forces have also contributed substantially to China’s
transition (Poston and
Gu 1987).
In most countries, there is an inverse relationship between TFR
and socioeco-
nomic development, with fertility declining as development
progresses (Bongaarts
and Watkins 1996; Bryant 2007). This is demonstrated in the
long-established TFR
differential between urban and rural areas (Jaffe 1942). Given
that urbanites tend to
have/prefer fewer children than rural residents, the process of
urbanization propels a
reduction of national TFR. Over two decades ago, Zeng and
Vaupel (1989)
observed that this process would likely decrease future birth
rates in China. At that
time, the urban–rural fertility differential remained quite large.

34. In 1986, the urban
TFR was 1.96 compared to the rural TFR of 2.72. Since China
was predominantly a
rural, agricultural society in the mid-1980s, Zeng and Vaupel
anticipated that the
national TFR had much room to decline through rural-to-urban
migration and the re-
classification of rural areas into urban areas. This would occur
as former rural
residents voluntarily adopted the preference for fewer children
that is prevalent
among urbanites or were compelled to have fewer children
because of the stricter
enforcement of the one-child policy in urban areas.
At the time Zeng and Vaupel made this observation, China’s
TFR was above the
replacement level and almost two-thirds of the population lived
in rural areas. At
present, China’s TFR is 1.47 (see Fig. 1) and over 46% of the
population resides in
urban areas (United Nations 2010). The proportional size of
China’s urban
population is below the global average (50%) and far below the
average (75%) for

35. developed countries. Hence, the potential for urban growth is
large and it is
expected that 73% of the Chinese population will live in urban
areas in 2050. What
is uncertain is how much urbanization can contribute to future
reductions in China’s
TFR. In general, our knowledge is limited about the
determinants of fertility
behavior in countries that are undergoing the process of
development but have low
fertility (Bongaarts 2002). This leaves questions about the
relationship between TFR
and urbanization in China, which cannot be considered a
developed country, but has
achieved sub-replacement fertility.
In China, the fertility differential between rural and urban areas
has narrowed
since 1978, but it is still large (see Fig. 2). In 2008, the TFR
was 1.73 in rural areas
and 1.22 in urban areas. The rural–urban TFR differential has,
moreover, remained
fairly stable since the early 1990s. If urban fertility behavior
remains consistent, this

36. 418 Z. Guo et al.
123
implies that urban expansion will propel further reductions in
China’s fertility.
According to Bongaarts (2002), at high levels of development
the relationship
between TFR and socioeconomic indicators is likely to be
nonlinear, because it is
unreasonable to expect an indefinite decline in TFR as
socioeconomic development
progresses. That is, although socioeconomic development
corresponds to a
reduction in fertility, it cannot totally extinguish the desire for
children. For China,
this relationship could become nonlinear at a comparatively
lower stage of
development as China’s TFR is already among the lowest in the
world. The
continuing urbanization of China appears to be inevitable, but it
is likely that at
some point this process will no longer lead to further reductions
in TFR.

37. The issue here is whether the relationship between TFR and
urbanization is
weakening in the Chinese context. The demographic trends
suggest this is the case.
0
0.1
0.2
0.3
0.4
0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
1978 1981 1984 1987 1990 1993 1996 1999 2002 2005 2008
T
F

38. R
TFR Urbanization rate
Year
D
egree of U
rbanization
Fig. 1 Trends of TFRs and urbanization rates, 1978–2008,
China. Source: National Bureau of Statistics
of China (2009)
0.5
1
1.5
2
2.5
3
3.5
1978 1981 1984 1987 1990 1993 1996 1999 2002 2005 2008
T
F
R

39. Rural
Urban
Year
Fig. 2 Trends of TFRs in urban and rural areas in 1978–2008.
Sources: The 1978–1987 data came from
the 2/1,000 fertility and contraceptive use survey conducted in
1988; the 1988–1992 data came from the
National Fertility Survey conducted in 1992; the 1993–2000
data came from the 2001 National Fertility
and Reproductive Health Survey; and the 2001–2008 data came
from the Annual Population Monitoring
Surveys
The Effect of Urbanization on China’s Fertility 419
123
Figure 1 illustrates that urbanization corresponded with a large
decline in TFR from
1978 and sometime in the mid-1990s. However, the TFR
plateaued thereafter, even
though the urbanization rate kept increasing. This raises the
question of how much
fertility changes (or can change) in response to urbanization in
low-fertility regimes.
To address this question, this study uses decomposition models

40. to assess the
contribution of urbanization to the decline in China’s TFR since
1978. The study
examines both the national and provincial levels because there
are regional
differences in socioeconomic development and the enforcement
of national birth
planning policies. In addition, the study simulates how much
more urbanization can
be expected contribute to fertility decline, under several
alternative scenarios of
urban expansion and birth planning reforms, forecasting TFR
until 2030.
Background
A key debate in the literature regards the primary source of
China’s fertility decline.
The debate is about how much socioeconomic factors have
contributed to this
decline, given the Chinese government’s tight regulation of
fertility behavior. The
predominant notion is that birth planning policies are the
fundamental reason for
China’s demographic transition (Poston and Gu 1987). These
interventions—which

41. reportedly have prevented over 300 million births since 1978
(Peng 2004)—have
led to doubts about whether the socioeconomic indicators that
were instrumental to
fertility decline in Western countries are also good explanations
for China’s
demographic transition. In China, the congruence of the timing
of fertility change
with the implementation of birth control policies is clear
evidence for the
importance of government intervention (Feeney and Wang
1993). For this reason,
the Chinese transition is considered to be a unique case among
the countries that
have reached sub-replacement levels of fertility (Cai 2010).
Though best known is the controversial one-child policy,
China’s efforts to
control population growth began well before this. The earliest
interventions came as
a response to the Great Famine of 1959–1961. After this natural
disaster, the
government began to set official targets for population growth
and provide better

42. access to birth planning information (Wu et al. 2009). However,
birth control did
not become a core aspect of economic planning until the 1970s
(Sharping 2003).
This started with the Wan-Xi-Shao (later-longer-fewer)
campaign, which promoted
later marriage and childbirth, longer birth intervals, and fewer
births (Liang and Lee
2006). Coinciding with the economic reforms, the Deng
Xiaoping administration
implemented the one-child policy in 1978 to improve China’s
prospects for
modernization and industrialization and to address concerns
about foodgrain
shortages (Wu et al. 2009). The one-child per couple rule
applies to around two-
thirds of Chinese couples, with most concessions to this rule
applying to couples
residing in rural areas (Cai 2010).
Feeney and Wang (1993) suggest that over one-half of China’s
fertility decline is
attributable to state intervention. There is no doubt that
government policies hastened
the ‘‘diffusion’’ of low fertility throughout China. In 1975, the
TFR was lower than

43. what could be expected from the level of development at that
time, and reflected the
success of the Wan-Xi-Shao campaign (Cai 2010). But state
intervention, and the
420 Z. Guo et al.
123
one-child policy in particular, is not the sole reason for the
decline in fertility. The
persistence of sub-regional variation in fertility after the
intensification of state
interventions appears to parallel sub-regional differences in
socioeconomic devel-
opment (Tien 1984). Cai (2010) observes that China’s fertility
in 2005 fell within a
range that could be expected from its level of development. In
addition, the TFR
remained above replacement levels during the 1980s, when the
one-child policy was
enforced with fewer exceptions than later. Cai concludes that
socioeconomic
development, in conjunction with state intervention, generated
an ideational shift

44. toward a preference for smaller families. For the one-child
policy to be the sole or
decisive factor, the fertility behavior of Chinese couples would
need to have been
radically different from the fertility behavior of couples in other
countries.
To some extent, China’s path to low fertility supports the
assumptions of
demographic transition theory (DTT). A central theme of DTT
is that the shift from
rural (agricultural) to urban (industrial) life initiates a change in
the economics of
childbearing (Kirk 1996). According to Notestein’s (1953)
classic argument, fertility
is high in agrarian societies as insurance against high mortality
and because children
were an important source of agricultural labor. Modernization
first leads to a reduction
in mortality, which decreases the need for high fertility to
insure population survival
(Bongaarts and Watkins 1996). The transition to industrial
economies (and urban
environments) also decreases the economic contributions of
children, whereas the

45. costs of their upbringing and education increase. Though no two
transitions are alike,
it remains plausible that modernization is responsible for
decreasing the need and
incentives for large families in numerous societies (Kirk 1996).
This theory of fertility change has been criticized for over-
emphasizing the role
of economic motivation (Hirschman 1994). To be sure, the
precise reasons for the
relationship between TFR and socioeconomic development are
difficult to ascertain,
and are surely irreducible to economic factors. Even though
DTT offers an
incomplete explanation of fertility change, this does not
undermine the empirical
relationship between TFR and levels of socioeconomic
development (Bryant 2007;
Cai 2010). The main criticism of DTT is not the relationship
between modernization
and fertility per se, but the mechanisms that constitute this
relationship (Bongaarts
and Watkins 1996). The criticism of DTT also focuses largely
on the role of

46. socioeconomic indicators in the onset and early phase of the
transition to low
fertility. However, Bongaarts (2002) observes that fertility
behavior is more
consistent with DTT at later stages of the transition, which is
our concern.
Of course, the relationship between fertility and socioeconomic
development
cannot be reduced to rational decisions about the costs/benefits
of children
(Hirschman 1994). However, DTT does not preclude other
causal variables and
indeed acknowledges the importance of ideational factors.
Notestein (1953)
observed that it is ‘‘impossible to be precise’’ about the
mechanisms that drive
fertility change in modern societies, and he indicated that
economic factors cannot
provide a sufficient explanation. He remarked that the
anonymity of urban life
weakened social control over fertility behavior and
modernization created more
opportunities for women outside the domestic sphere.
Urbanization is a proxy for

47. changes in social norms and gender roles, which, together with
economic forces,
generate a preference for smaller families.
The Effect of Urbanization on China’s Fertility 421
123
The economics of children and ideational preferences for
smaller families are
important components of the relationship between fertility
decline and socioeco-
nomic development in China (Cai 2010). However, rural–urban
differences in the
enforcement of the one-child policy suggest that an increasing
proportion of urban
residents will lead to an inevitable decline in national fertility,
unless China reforms
the policy. The one-child rule is strictly enforced in all urban
areas in China and
throughout 6 provinces (Gu et al. 2007). There are some
exceptions for couples that
have agricultural household registration status. In 19 provinces
rural couples are
allowed a second child if their first child is a girl and in another

48. 5 provinces all rural
couples are permitted two children. The urban population
remained stable until
1978, but the relaxation of official restrictions on rural-to-urban
migration and the
reclassification of rural areas into urban areas have fueled the
proliferation of the
urban population (Zeng and Vaupel 1989). This process is
exposing a growing
number of Chinese to urban values and subjecting them to the
one-child rule.
Methods
The data for the country-level TFR and the proportion of urban
females of
reproductive age come from the 1982 Census, the 1990 Census,
and the 2001 and
2008 one per thousand population surveys conducted by the
National Bureau of
Statistics of China (NBS 2009). The estimates for provincial-
level fertility are
drawn from an NBS and East–West Center (2007) report. A
decomposition
approach is used to model the effects of urbanization on fertility
change. Following

49. Das Gupta (1991), the analysis decomposes TFRasfr into three
components to
estimate the separate effects of changes in urban fertility, rural
fertility, and
urbanization on TFR. For the reader’s convenience, the
mathematical expression is
recapitulated as below.
TFR can be formulated as TFRasfr ¼ 5
P
x Fx, where Fx is the age-specific birth
rate for the 5-year age group starting at age x. Fx can be
expressed as a weighted
sum of urban-age-specific birth rate(Fx,u) and rural-age-
specific birth rate (Fx,r),
where the weights kx,r and kx,u are the proportion of women in
age group x to x?5
residing in rural and urban areas, respectively (here we have
kx,r ? kx,u = 1,
Dkx,u = -Dkx,r). This leads to the reformulation of TFRasfr,
TFRasfr ¼ 5RxFx ¼ 5Rx Fx;rkx;r þ Fx;ukx;u
� �
ð1Þ
It follows that the change in the TFRasfr is,
DTFRasfr ¼ 5Rx Fx;u � Fx;r
� �
Dkx;u þ 5Rxkx;rDFx;r þ 5RFx;u DFx;u ð2Þ

50. where the symbol D denotes change, and Fx;r, Fx;u, kx;r and
kx;u are average values
over the period. The first of the three principal terms on the
right hand side of
Eq. 2 denotes the contribution to change in TFR from changes
of the age-specific
proportion of urban females within the total female population
at reproductive
age. The second term denotes the contribution from changes in
age-specific rural
fertility. The third term denotes the contribution from changes
in age-specific of
urban fertility.
422 Z. Guo et al.
123
To demonstrate the results of this decomposition exercise, we
begin with
the scenario where (i) the rural fertility is always higher than
urban fertility in any
age-group, (Fx,r [ Fx,u), (ii) all components have no changes
during the period
(Dkx,u = DFx,r = DFx,u = 0) This situation is illustrated in Fig.
3a, in which all

51. TFRs are constant during the period.
Suppose now that, under the same assumptions (i and ii), we
now allow the proportion
of urban females to increase at each reproductive age (Dkx,u [
0). Equation 2 is
simplified: DTFRasfr ¼ 5Rx Fx;u � Fx;r
� �
Dkx;u where Fx;u � Fx;r0 according to
assumption (i) and DTFRasfr 0. As shown Fig. 3b, this change
drives down the
national fertility despite that both urban and rural fertility
remains unchanged.
Furthermore, if there is a positive change in both urban and
rural age-specific birth rate
(DFx,u [ 0 and DFx,r [ 0) in the proportion of urban females
(Dkx,u [ 0), the first term
of Eq. 2 becomes negative but the second and third terms turn
positive, such that the
change of national TFR, as the sum of the three terms, can be
unchanged
(DTFRasfr = 0). This scenario is demonstrated in Fig. 3c. In
short, this illustration
demonstrates that the trends of national fertility, urban fertility
and rural fertility may
not be in the same direction when we take into account the role
of urbanization.

52. National TFR, 1982–2008
Table 1 presents the change in age-specific national TFR, which
is decomposed into
three components. The first component represents changes in
rural fertility, the
TFR
Time
TFR(rural)
TFR(nation)
TFR(urban)
TFR
Time
TFR(rural)
TFR(nation)
TFR
Time
TFR(rural)
TFR(nation)
(a) (b)

53. (c)
TFR(urban)
TFR(urban)
Fig. 3 Illustrations of urbanization effects on fertility
The Effect of Urbanization on China’s Fertility 423
123
second component represents changes in urban fertility, and the
third component
represents the influence of urbanization, i.e., changes in the
proportion of urban
females aged 15–49 years. The results show a 1.15 decrease in
China’s TFR
between 1982 and 2008. The change in rural fertility behavior
contributed to 0.83 of
this decrease and the change in urban fertility behavior
contributed to 0.07 of this
decrease. The change of urbanization resulted in a 0.25 decrease
in the national
TFR, which represents about 22% of the total reduction in TFR
from 1982 to 2008.
The change in rural fertility behavior accounted for the largest

54. amount (72%) of the
decline in TFR during this period.
From 1982 to 1990, both the changes in rural fertility behavior
and urbanization
led to a reduction in national TFR. During this period, there was
an increase in
urban births, and thus urban fertility behavior had a positive
impact on national
TFR. The reduction of national TFR through rural fertility
behavior and
urbanization likely reflect the impact of the one-child policy.
The results suggest
a tightening up of the one-child policy in rural areas, such as
preventing 3rd and
higher order births. The effect of urbanization is presumably a
result of a greater
number of people becoming adherents to the strict one-child
rule through permanent
migration or the reclassification of rural areas into urban areas.
The impact of urban
fertility behavior is not that surprising. Urban fertility has been
considerably lower
than rural fertility since the 1960s and it reached the sub-
replacement level in the

55. early 1970s (Zeng and Vaupel 1989). Given that the urban TFR
was 1.4 in 1981
(Fig. 2), it is unreasonable to anticipate that it could decline
much more.
Changes in rural fertility behavior, urban fertility behavior, and
levels of
urbanization all contributed to the reduction in national TFR
from 2.30 in 1990 to
1.39 in 2001. The change in rural fertility behavior had the
greatest effect,
accounting for 69% of the decline in national TFR. The effect
of urban fertility
behavior accounted for 20% of the decline in TFR and
urbanization accounted for
the remaining 11%. From 2001 to 2008, the national TFR
increased from 1.39 to
1.47. This was a result of growth in both rural and urban
fertilities. However, the
effect of urbanization on national TFR growth was negative.
The rebound of
national TFR demonstrates the challenge of reducing TFR in
low-fertility regimes.
But these results also show that urbanization remains a source
of declines in TFR in

56. developing countries with low fertility.
Table 1 demonstrates that urbanization was primary reason for
the decline in
China’s TFR between 2001 and 2008. To illustrate the
independent effect of
urbanization on fertility change, we compared the national TFR
with and without
Table 1 Decomposition of the Changes in TFR in China: 1982–
2008
Period TFR (per 1,000) Absolute change (per 1,000)
Start End Change Rural Urban Urbanization
1982–1990 2,620 2,300 -320 -310 60 -70
1990–2001 2,300 1,390 -910 -630 -180 -100
2001–2008 1,390 1,470 80 110 50 -80
1982–2008 2,620 1,470 -1150 -830 -70 -250
424 Z. Guo et al.
123
the effect of urbanization. In Fig. 4, the dotted line represents
what the national TFR

57. would be without urbanization (counter-factual test). This
figure confirms the
importance of urbanization to the decline in China’s fertility.
Without urbanization,
China’s TFR would be higher than it actually is.
Province-Level TFR in 2000 and 2005
In this section, we present the decomposition of changes in TFR
for 30 of 31
provinces and municipalities in mainland China. The region of
Tibet is excluded
because the sample size of birth numbers is too small to permit
an accurate analysis.
In China, socioeconomic development has been uneven and
there are disparities
between the provinces (Peng 2011). The national results
presented above could,
therefore, provide an incomplete picture of the relationship
between TFR and
urbanization. China’s provinces and municipalities fall under
four levels of
urbanization (Fu et al. 2009). The first level includes
municipalities such as
Shanghai, Beijing, and Tianjin, which are China’s economic
powerhouses and have

58. the highest national levels of urbanization. The second are
provinces at a medium
level of urbanization, including Heilongjiang, Jilin, and
Liaoning. The third level
consists of nine provinces with low levels of urbanization:
Guangdong, Jiangsu,
Shandong, Hubei, Shanxi, Qinghai, Xinjiang, Hainan, and
Ningxia. The remaining
provinces have very low levels of urbanization.
Figure 5 plots the province-level TFRs according to degree of
urbanization. This
figure indicates that there is, in general, an inverse relationship
between TFR and
urbanization. In accordance, the most urbanized provinces also
had the lowest TFRs
in 2000 and 2005. However, there are incidences where low
levels of urbanization
are associated with high levels of fertility. Table 2 provides
additional evidence for
1
1.1
1.2

59. 1.3
1.4
1.5
1.6
1.7
1.8
2001 2002 2003 2004 2005 2006 2007 2008
T
F
R
Year
Nation
Counter-factual
Urban
Rural
Fig. 4 Counter-factual test on TFRs from 2001 to 2008. Source:
National Bureau of Statistics of China
(2009)
The Effect of Urbanization on China’s Fertility 425
123

60. this relationship. Between 2000 and 2005, both changes in
urban fertility behaviors
and levels of urbanization contributed to a decrease in the
national TFR, but these
effects were somewhat offset because of an increase in the rural
TFR across China.
During this time, the TFR declined in 23
provinces/municipalities. The greatest
decreases occurred in the four large metropolitan
municipalities, Beijing, Tianjin,
Shanghai, and Chongqing. The majority of reduction in these
provincial TFRs is
attributable to changes in urban fertility behavior and the
expansion of the urban
population. Other provinces also experienced large reductions
in their TFRs. These
include three coastal provinces, Liaoning in the north and
Guangdong and Hainan in
the south, and three inland provinces, Shanxi, Henan, and
Jiangxi.
However, the relationship between TFR and socioeconomic
development is not

61. entirely consistent across China. Several lesser developed
provinces (Guizhou,
Yunnan, Qinghai, and Xinjiang) also experienced large declines
in their TFRs. The
declines in these provinces were largely a result of changes in
fertility behavior in rural
areas. In six inland provinces (Guangxi, Sichuan, Hubei,
Jiangsu, Hebei, and Anhui)
the TFR increased. In some of these provinces, the relationship
between TFR and
urbanization does not appear to be as robust as it is elsewhere,
but this is generally
because high fertility in rural areas offset the effect of
urbanization. Shandong is the
only coastal province that experienced a large increase in its
TFR. That said, Fig. 6
illustrates that, between 2000 and 2005, increases in levels of
urbanization associated
with a decline in the TFRs in all provinces except for Jilin,
Shanghai, and Xinjiang.
The decompositions presented in Table 2 suggest that changes
in urban fertility
behavior in the most urbanized provinces accounted for most of
reduction in TFR

62. observed among them. These provinces (and municipalities) are
Beijing, Tianjin,
Shanghai, Liaoning, Jilin, Heilongjiang, Zhejiang, and
Guangdong. In these
provinces, an average of 65% of women aged 15–49 reside in
urban areas,
compared to the national average of 41%. The declines in the
number of urban
500
1000
1500
2000
2500
0.2 0.4 0.6 0.8 1.0
T
F
R
(
pe
r
1,
00

63. 0)
Degree of urbanization
2000 2005 2000 2005
Fig. 5 Trends of urbanization rates and TFRs in 2000 and 2005.
Sources: ‘‘Fertility estimates for
provinces of China’’ (National Bureau of Statistics and the
East–West Center 2007) and The report of
China’s 2005 national 1% population survey. Beijing: National
Bureau of Statistics of China 2006
426 Z. Guo et al.
123
births in these provinces represented an important source of the
decrease in the
national TFR. While changes in rural fertility behavior
contributed much to decline
in the national TFR from 1982 to 1990 and also from 1990 to
2001 (Table 1), this
effect seems to have ebbed in recent years. Between 2000 and
2005, rural fertility
had a positive effect on the national TFR, even though this
effect was offset because
of decreases related to urban fertility behavior and urbanization.

64. Table 2 Decomposition of the changes in TFR: Chinese
Provinces, 2000–2005
Absolute change in TFR (per 1,000) Relative change in TFR
(percent)
Total Rural Urban Urbanization Rural Urban Urbanization
China -74 23 -51 -45 1.6 -3.7 -3.2
Beijing -202 2 -189 -16 0.2 -21.7 -1.8
Tianjin -177 -3 -146 -27 -0.3 -14.8 -2.7
Hebei 101 142 -6 -36 9.7 -0.4 -2.4
Shanxi -173 -85 -72 -15 -5.2 -4.4 -0.9
Neimenggu -27 -4 3 -26 -0.3 0.3 -2.2
Liaoning -104 21 -98 -28 1.9 -9 -2.5
Jilin -13 54 -69 2 5.4 -7 0.2
Heilongjiang -12 52 -63 -1 5.2 -6.2 -0.1
Shanghai -389 -20 -371 2 -1.9 -34.8 0.2
Jiangsu 52 152 -49 -52 13.6 -4.4 -4.6
Zhejiang -69 37 -79 -27 2.8 -6 -2.1
Anhui 301 323 61 -83 21.8 4.1 -5.6
Fujian -29 104 -80 -53 8.5 -6.5 -4.4

65. Jiangxi -120 -70 18 -68 -3.9 1 -3.8
Shandong 253 204 84 -35 16.1 6.6 -2.7
Henan -301 -221 -44 -36 -13.9 -2.8 -2.2
Hubei 119 104 41 -25 8.3 3.3 -2
Hunan -16 13 11 -40 0.9 0.7 -2.7
Guangdong -421 -78 -313 -30 -5.4 -22 -2.1
Guangxi 43 88 2 -47 5 0.1 -2.7
Hainan -226 -56 -102 -68 -3.1 -5.5 -3.7
Chongqin -239 -36 -92 -112 -2.5 -6.4 -7.8
Sichuan 36 92 -30 -27 6.3 -2 -1.8
Guizhou -450 -367 -50 -32 -15.4 -2.1 -1.3
Yunnan -307 -240 -19 -49 -11.8 -0.9 -2.4
Shaanxi -85 -11 -51 -23 -0.9 -4 -1.8
Gansu -29 -7 -13 -10 -0.5 -0.9 -0.6
Qinghai -435 -275 -138 -22 -15.1 -7.5 -1.2
Ningxia -19 30 -4 -45 1.7 -0.2 -2.6
Xinjiang -156 -109 -61 14 -6.3 -3.5 0.8
Sources: Fertility estimates for Provinces of China National
Bureau of Statistics and the East–West

66. Center (2007) and The report of China’s 2005 national 1%
sample survey. Beijing: National Bureau of
Statistics of China 2006
The Effect of Urbanization on China’s Fertility 427
123
In 12 provinces, change in rural fertility behavior was
instrumental in propelling
either the growth or the reduction of provincial-level fertility
between 2000 and
2005. These 12 provinces can be classified has having
comparatively small urban
populations. In seven of these provinces (Hebei, Jiangsu, Anhui,
Shandong, Hubei,
Guangxi, and Sichuan) the provincial-level TFR increased
because of increases of
fertility in rural areas. In several of these provinces the TFR
increased despite a
decrease in urban fertility and a negative effect of urbanization.
Moreover,
urbanization had a negative effect on the TFR in each of these
provinces, and
fertility in urban areas increased in only in Anhui and Guangxi.

67. In some provinces,
such as Henan, Guizhou, and Yunnan, the reduction in their
TFRs was mainly a
result of declines of fertility in rural areas.
Future Effects of Urbanization
The evidence presented above suggests that urbanization is an
important factor in
the reduction of China’s TFR. The question that remains is
whether urbanization
will have a negative effect on China’s fertility in the future. To
address this
question, we forecasted China’s fertility from 2010 to 2030,
using six scenarios
based on three assumptions about urban growth and two
assumptions about
differences in rural and urban fertilities. Under our low-growth
assumption, 62% of
the population will be urban in 2030. In the medium-growth
assumption, the
proportion of the urban population will be 67% in 2030. In the
high-growth
assumption, the urban population will account for 84% of the
general population in

68. -8 -7 -6 -5 -4 -3 -2 -1 0 1
Chongqin
Anhui
Jiangsu
Fujian
Jiangxi
Hainan
Hunan
Guangxi
Shandong
Tianjin
Ningxia
Liaoning
Yunnan
Hebei
Henan
Neimenggu
Guangdong
Zhejiang
Hubei
Sichuan
Beijing
Shaanxi
Guizhou
Qinghai
Shanxi
Gansu
Heilongjiang
Jilin
Shanghai
Xinjiang
Percent changeFig. 6 Effect of change in the
urbanization rate on TFR in
2000–2005, Chinese Provinces

69. 428 Z. Guo et al.
123
2030. The figures for the medium-growth scenario best accord
with official
estimates of future urbanization (Pan and Wei 2010). Because
predicted data are not
age-specific based, a simplified version of the decomposition
equation is introduced
and presented in Appendix A.
We considered these three assumptions about urban growth
under two different
assumptions about future differences in rural and urban
fertilities. First, we used a
time series model to project the stochastic pattern of rural and
urban fertilities.
Details about the stochastic model are presented in Appendix B.
In this model, rural
TFR is stable at 1.6 and urban TFR is stable at around 1.1, for a
fairly persistent
difference of 0.5 between them. Second, we used a model of the
rural–urban TFR

70. differential that assumes that the birth planning policy has been
relaxed to a two-
child rule for all couples. Under this assumption, the rural TFR
would be 1.88 and
the urban TFR 1.5 in and after 2010 (see Zheng 2004). While
the second assumption
suggests a narrowing gap of rural and urban TFRs (0.38), it is
unreasonable to
expect rural and urban fertility behaviors will converge in the
next 20 years, even if
the one-child policy is relaxed in urban areas.
Table 3 presents the estimated TFRs under these six scenarios
of urbanization
and differences in rural and urban fertility. In all six scenarios,
urbanization is
projected to be the primary factor behind fertility change from
2010 to 2030, and the
national TFR will remain at sub-replacement levels. The
stochastic projections
result in little variation in rural and urban fertilities during this
time. Changes in
rural and urban fertility behaviors are projected to have small
negative effects on
China’s TFR under the present birth planning policy. Under the

71. ‘‘relaxed policy’’
assumption, rural and urban fertility do not affect the national
TFR. This implies
that the projected decline in TFR will occur entirely through
urbanization. Under
medium-growth (the expected level of urbanization), the
national TFR will decrease
from 1.44 in 2010 to 1.25 in 2030 under the stochastic
assumption and from 1.7 to
1.6 under the ‘‘relaxed policy’’ assumption.
Conclusions
China has experienced rapid urbanization since 1978 and the
urban population is
projected to continue growing for several more decades. As
noted above, there is an
Table 3 Decomposition of the predicted TFRs: 2010–2030
Urbanization
development
Fertility
assumption
TFR (per 1,000) Absolute change (per 1,000)

72. Start End Change Rural Urban Urbanization
High growth ‘‘Stochastic’’ 1,440 1,200 -240 -20 -30 -190
‘‘Relaxed’’ 1,700 1,560 -140 0 0 -140
Medium growth ‘‘Stochastic’’ 1,440 1,250 -190 -20 -30 -140
‘‘Relaxed’’ 1,700 1,600 -100 0 0 -100
Low growth ‘‘Stochastic’’ 1,440 1,310 -130 -10 -40 -80
‘‘Relaxed’’ 1,700 1,640 -60 0 0 -60
The Effect of Urbanization on China’s Fertility 429
123
inverse relationship between TFR and urbanization. This study
examined the effects
of urbanization on fertility change in China between 1978 and
2008, and projected
how much more urbanization can be expected to contribute to
fertility change
between 2010 and 2030. This study decomposed China’s present
and future TFR
into three components to estimate the separate effects of
changes: the effect of
change in rural fertility behavior, the effect of change in urban

73. fertility behavior,
and the effect of urbanization. The study assumed that regional
differences in levels
of urbanization could influence the relationship between
national TFR and
urbanization. Hence, the analysis includes findings for the
decomposed effects on
the national and provincial-level TFRs for 2000–2005.
The study offers three major conclusions about past and future
fertility trends.
First, the change in rural fertility behavior accounted for most
of the decline in the
national TFR from 1982 to 2008. The national TFR declined
from 2.62 to 1.47
during this period. The reduction in rural fertility was
responsible for 72% of this
decline. This finding suggests that the one-child policy was the
primary instrument
of China’s achievement of sub-replacement fertility. Between
2000 and 2005,
several less developed provinces (e.g., Guizhou, Yunnan,
Xinjiang) experienced
large declines in their TFRs largely because of reductions in the
number of rural

74. births. It is possible that some of these declines in rural fertility
is related to other
aspects of socioeconomic development, such as improvements
in the educational
attainment of rural residents or decreases in need for
agricultural labor, but the one-
child policy is likely the main factor for this change. In seven
provinces, however,
an increase of fertility in rural areas was the driving factor for
increases in province-
level TFR, which could reflect local variation in the
enforcement of the one-child
policy. As the majority of Chinese (54%) still live in rural
areas, it is unsurprising
that this population remains the vanguard of China’s fertility
transition.
Second, the contribution of urbanization to the decline of
China’s TFR between
1982 and 2008 was modest in comparison to the large effect that
decreases in rural
fertility had. However, urbanization was indeed an important
factor and it had a
negative effect on the national TFR in each of the periods
observed (1982–1990,

75. 1990–2001, 2001–2008, and 1982–2008). About 22% of the
reduction in the
national TFR between 1982 and 2008 is related to the process of
urbanization.
Moreover, the findings suggest that urbanization has recently
become the principal
source for curbing population growth. From 2001 to 2008,
urbanization had a
negative effect on the national TFR, but increases in rural and
urban births offset
this effect. In all but three provinces, urbanization was
associated with a decline in
province-level TFRs between 2000 and 2005. The three
exceptions (Jilin, Shanghai,
and Xinjiang) had relatively low rates of urbanization during
this period, thus the
impact of urbanization on TFR in these areas was also minimal.
Low rates of
urbanization and possible measurement errors in TFRs may
explain the unexpected
relationship between TFR and urbanization among them. In
contrast, in provinces
with high rates of urbanization and large rural–urban fertility
differentials, the effect

76. of urbanization on province-level TFR is quite pronounced.
Given the short period of observation for changes in province-
level TFRs
(5 years), it is possible that the findings presented here do not
reflect the full effect
of urbanization. The intent here is to disentangle the effect of
urbanization from the
430 Z. Guo et al.
123
effect of urban fertility behavior. To some extent, the change in
urban fertility
behavior is likely a ‘‘lagged’’ effect of urbanization. That is,
the effects of rural-to-
urban migration and the reclassification of rural areas into
urban areas on urban
fertility are not immediate. Rather, these new urbanities
gradually adopt urban
fertility behaviors and are exempt from the strict one-child rule
in the short-term.
This effect is difficult to decompose because of data limitations,
but it suggests that

77. a portion of the decreases in the national and province-level
TFRs related to changes
in urban fertility behavior represent an unobserved effect of
urbanization.
Finally, the findings suggest that urbanization will take over as
the main engine
of fertility decline from 2010 to 2030. This is evident from
recent trends. While the
national TFR increased from 1.39 to 1.47 from 2001 to 2008,
this change would
have been larger without the negative effect of urbanization.
After several decades
of birth planning, it appears that the one-child policy is
reaching the limits of what it
can accomplish. Our projections indicate that changes in rural
and urban fertility
behaviors have small effects on the TFR under the current
policy. In general, it is
becoming increasingly difficult to decrease the TFR, given that
it is already very
low. However, given the levels of urbanization that can be
realistically expected in
2030 and beyond, relaxing the one-child policy to a two-child
policy would not have

78. a major effect on China’s population growth. Under this
scenario, we project the
TFR to be 1.6 in 2030. This supports studies that call for
alternative policies to the
one-child rule (e.g., Greenhalgh and Bongaarts 1987; Wang
2005; Zeng 2007).
Acknowledgments This study is jointly supported by Social
Sciences and Humanities Research
Council of Canada, Program for Chang Jiang Scholars and
Innovative Research Team in Universities of
the Ministry of Education of China (IRT0855) and the National
985 Project of the Ministry of Education
and Treasury Department of China (07200701). The authors
gratefully acknowledge helpful comments
from Barry Edmonston.
Appendix A: A Simplified Version of the Decomposition
Equation
To derive a simplified version of decomposition equation, Eq. 1
requires three
additional assumptions: (a) Fx is constant for all x, i.e., age
specific fertility rates are
constant in all ages; (b) kx is constant for all x, meaning that
the proportion of urban
females at aged x in the total female population is constant; and
(c) the sex
composition of urbanites remains constant while urbanization
rate (Cu) increases

79. such that Cu = ku. Under these assumptions, Eq. 1 can be re-
written as
TFRasfr ¼ FrCr þ FuCu ð3Þ
and decomposing (3),
DTFR ¼ Fu � Fr
� �
DCu þ CrDFr þ CuDFu ð4Þ
where Fr and Fu denote rural TFR and urban TFR, respectively;
Cr and Cu denote the
proportion of rural and urban population; and again we have Cr
? Cu = 1. Table 4
shows that the difference in TFR between using Eqs. 1 and 3 is
minimal (see the last
column of Table 4), suggesting that it is not unreasonable to
decompose TFR, rather
than TFRasfr, in the decomposition exercise and the forecasts of
TFRs (see Appendix B).
The Effect of Urbanization on China’s Fertility 431
123
Appendix B: Forecast of TFRs in Urban and Rural Areas
To forecast future national fertility, we estimated a
conventional time series model

80. for the log-transformed rural and urban TFRs (Fr and Fu),
conditional upon that the
TFRs are greater than 0 (e.g., Box et al. 2008) We used data
from 1950 to 2008. The
fitted models for Fu and Fr are given below (standard errors in
parentheses):
ln Fu;t ¼ 0:992 0:0096ð Þ � ln Fu;t�1; R2 ¼ 0:904
ln Fr;t ¼ 0:995 0:0065ð Þ � ln Fr;t�1;R2 ¼ 0:903
Using these equations, it is straightforward to forecast TFRs in
urban and rural area
for the next 20 years (see Fig. 7). Figure 7 shows that rural
TFRs in next 20 years
are fairly stable at approximately 1.6, while urban TFRs are
around 1.1.
Table 4 A simplified decomposition of TFRs
Year TFR (rural) TFR (urban) Urbanization rate (%) TFR (1)
TFR (3) Difference
1982 3.02 1.40 21.13 2.62 2.68 0.06
1990 2.58 1.59 26.41 2.30 2.32 0.02
2001 1.60 1.08 37.66 1.39 1.40 0.01
2008 1.73 1.22 45.68 1.47 1.50 0.03
Sources: The 1982 Census, the 1990 Census; the 2001 and 2008
0.1% population surveys

81. 0
1
2
3
4
5
6
7
8
1950 1960 1970 1980 1990 2000 2010 2020 2030
T
F
R
Year
TFR(rural)
TFR(urban)
Fig. 7 Forecast of TFRs in urban and rural areas. Sources: The
1950–1977 data came from 1/1,000
fertility survey conducted in 1982; the 1978–1987 data were
from the 1988 2/1000 fertility and

82. contraceptive use survey; the 1988–1992 data came from the
1992 National Fertility Survey; the
1993–2000 data were obtained from the 2001 National Fertility
and Reproductive Health Survey; and
finally the 2001–2008 data came from the Annual Population
Monitoring Surveys
432 Z. Guo et al.
123
References
Bongaarts, J. (2002). The end of the fertility transition in the
developing world. Population Bulletin of the
United Nations, 48(49), 271–286.
Bongaarts, J., & Watkins, S. C. (1996). Social interactions and
contemporary fertility transitions.
Population and Development Review, 22(4), 639–682.
Box, E. P., Jenkins, G. M., & Reinsel, G. C. (2008). Time series
analysis: Forecasting and control. New
York: Wiley.
Bryant, J. (2007). Theories of fertility decline and evidence
from development indicators. Population and
Development Review, 33(1), 101–127.
Cai, Y. (2010). China’s below-replacement fertility:
Government policy or socioeconomic development?
Population and Development Review, 36(3), 419–440.

83. Das Gupta, P. (1991). Decomposition of the difference between
two rates and its consistency when more
than two populations are involved. Mathematical Population
Studies, 3(2), 105–125.
Feeney, G., & Wang, F. (1993). Parity progression and birth
intervals in China: The influence of policy in
hastening fertility decline. Population and Development
Review, 19(1), 61–101.
Feeney, G., & Yuan, J. (1994). Below replacement fertility in
China? A close look at recent evidence.
Population Studies, 48(3), 381–394.
Fu, Y., Wei, P., & Jin, R. (2009). Urbanization level in the
transitional stage of China and provincial
economic growth. Paper presented at the 2009 International
Conference on Management and Service
Science, Wuhan, China.
Greenhalgh, S., & Bongaarts, J. (1987). Fertility policy in
China: Future options. Science, 235(4793),
673–701.
Gu, B. (2007). Low fertility in China: Trends, policy, and
impact. Asia Pacific Population Journal, 22(2),
73–90.
Gu, B., Wang, F., Guo, Z., & Zhang, E. (2007). China’s local
and national fertility policies at the end of
the twentieth century. Population and Development Review,
33(1), 129–147.
Hirschman, C. (1994). Why fertility changes. Annual Review of
Sociology, 20, 203–233.

84. Jaffe, A. J. (1942). Urbanization and fertility. American Journal
of Sociology, 48(1), 48–60.
Kirk, D. (1996). Demographic transition theory. Population
Studies, 50(3), 361–387.
Liang, Q., & Lee, C. (2006). Fertility and population policy: An
overview. In D. L. Poston, C. Lee,
C. Chang, S. L. McKibben, & C. S. Walther (Eds.), Fertility,
family planning, and population policy
in China (pp. 8–19). New York: Routledge.
National Bureau of Statistics of China (NBS). (2009). Chinese
statistics yearbook, 2009. Beijing:
National Bureau of Statistics.
National Bureau of Statistics and the East-West Center. (2007).
Fertility estimates for the provinces of
China, 1975–2000. Beijing: National Bureau of Statistics.
Notestein, F. (1953). Economic problems of population change.
Proceedings of the eighth international
conference of agricultural economists (pp. 13–31). London:
Oxford University Press.
Pan, J. H., & Wei, H. K. (2010). Annual report on urban
development of China. Beijing: Academic Press
of Social Sciences.
Peng, X. (2004). Is it time for China to change its population
policy? China: An International Journal,
2(1), 135–149.
Peng, X. (2011). China’s demographic history and future
challenges. Science 333 (Special section),
581–587.

85. Poston, D. L., & Gu, B. (1987). Socioeconomic development,
family planning, and fertility in China.
Demography, 24(4), 531–551.
Sharping, T. (2003). Birth control in China 1949–2000:
Population policy and demographic
development. New York: Routledge Curzon.
Tien, H. Y. (1984). Induced fertility transition: Impact of
population planning and socio-economic change
in the People’s Republic of China. Population Studies, 38(3),
385–400.
United Nations. (2010). World urbanization prospects: The 2009
revision. New York: United Nations.
Wang, F. (2005). Can China afford to continue its one-child
policy? Asian Pacific Issues, 77, 1–12.
Wu, Z., Schimmele, C. M., & Li, S. (2009). Demographic
change and economic reform. In A. Sweetman
& J. Zhang (Eds.), Economic transitions with Chinese
characteristics: Social change during thirty
years of reform (pp. 149–167). Kingston: McGill-Queen’s
University Press.
The Effect of Urbanization on China’s Fertility 433
123
Zeng, Y. (2007). Options for fertility policy transition in China.
Population and Development Review,
33(2), 215–246.
Zeng, Y., & Vaupel, J. W. (1989). The impact of urbanization

86. and delayed childbearing on population
growth and aging in China. Population and Development
Review, 15(3), 425–445.
Zheng, Z. (2004). Fertility desire of married women in China.
Chinese Journal of Population Studies, 5,
73–78.
434 Z. Guo et al.
123
The Effect of Urbanization on China’s
FertilityAbstractIntroductionBackgroundMethodsNational TFR,
1982--2008Province-Level TFR in 2000 and 2005Future Effects
of UrbanizationConclusionsAcknowledgmentsAppendix A: A
Simplified Version of the Decomposition EquationAppendix B:
Forecast of TFRs in Urban and Rural AreasReferences